in PDF Format - Embedded-Control-Europe.com

Dear Readers, 

It is fact that standards swarm 

the market for embedded computing. 

And it s also fact that 

there are many so called standards 

that it needs some effort to find 

the right standard for a given application. 

Probably it started all 

with x86, the first standard coming 

from the PC world and will 

certainly not end with COM Express 

,one of these actual standards 

which gained great momentum 

and therefore importance 

in embedded computing 

launched in 2004. In the COM 

Express COM.0 specification, the 

PICMG defines the standard for a Computer-on-Module (COM) as 

a bootable host computer in the form of a single large-scale integrated 

component. The vendor-independent specification of interfaces and 

form factors for Computer-on-Modules gives designers and solution 

providers a firm basis on which to develop products for the market 

that are future-oriented and promise long-term availability. PICMG 

continues along the right path with the Rev 2.0 of the COM Express 

specification. You ll find the new features and differences between 

COM Express COM.0 specification and the actual Rev. 2.0 in an 

article at page 24. 

But back to x86 systems. Many of the standards in embedded computing 

are based on this primary standard but nevertheless there is 

still room for proprietary embedded computing systems. It is wellknown 

in the industry that standardization of x86 systems for 

industrial applications dates back to the nineties, when the first 

PC/104 systems and ISA slot cards were introduced. Standardization 

here refers to the form factor and the joint bus system, in this case the 

ISA bus. The advantage for the user was readily apparent in the large 

number of exchangeable or complementary systems by various 

vendors. Over the years, additional bus systems, such as the PCI bus 

and now PCIe bus, were added, leading to new standards. The most 

recent version of a corresponding system with PCI and PCIe bus is 

COM Express. 

If, however, the application is closer to a direct machine control, different 

interfaces and functions play a role. Serial interfaces, GPIOs 

and field buses are important here. In the PC world, these interfaces 

are implemented through in part elaborate additional cards. This is 

exactly where ARM-based processors apply - the combination of PCbased 

basic functions coupled with special machine-oriented interfaces. 

The significantly lower power loss is invoked by processor manufacturers 

as an additional criterion. The diversity of processor variants 

outside the x86-world for which module solutions and so-called standards 

are offered is almost limitless. An article starting at page 30 

describes that proprietary systems are useful in the non-x86 market 

since they provide the full functionality and strengths of the chosen 

processor on the one hand and, to a certain extent, offer the advantages 

of the interchangeability of x86 systems. 

Yours sincerly 

Wolfgang Patelay 

(Editor) 

VIEWPOINT 

3 September 2010

CONTENTS 

Viewpoint 3 

Defence & Aerospace 

ATCA is now ready for defence 

and aerospace applications 5 

CompactPCI cards for advanced 

avionics systems development 8 

Ruggedized VPX provides extra-robust 

systems for demanding applications 11 

Transportation 

Will Ethernet be the only train bus 

in the trains of the future? 13 

Industrial Automation 

Industrial imaging trends lead 

to FPGA-based GPUs and interfaces 16 

Modular controllers enable platform 

concept for custom applications 20 

Embedded designers learn their 

lessons from the battlefield 22 

Small Form Factor Boards 

PICMG adopts new specification 

for COM Express modules 24 

Replacing discontinued modules forces 

ETX module transplantation 28 

Can proprietary ARM-based systems 

exist alongside the x86 standards? 30 

Software Development 

Adding software security using 

virtualized solutions 33 

Software development optimised 

through unit testing 35 

AdvancedTCA 

Performance gains from multi-core 

processors partnered with ATCA 38 

Product News 41 

Cover Photo 

AMD 

September 2010 4 


and aerospace applications Page 5 

This article explains why ACTA is an ideal platform for system 

developers seeking rapid development at low cost. Further, the 

flexibility of the architecture gives ATCA the ability to meet many 

different military networking needs, helping ensure interoperability 

among diverse systems for information exchange and avoiding 

stovepipe installations. 


avionics systems development Page 8 

CompactPCI is a first choice for 

modernizing legacy test equipment 

or developing a new system. Its modular 

hardware design, wide range of 

COTS I/O options, good instrumentation 

control, and clear migration path 

to future technology make it a smart, 

cost-effective and flexible solution 

able to support avionics systems development for years to come. 


in the trains of the future? Page 13 

This article discusses the various train busses used today and the 

possibility that Ethernet could become the train bus of the future. 


concept for custom applications Page 20 

Despite the varied purposes that 

industrial controllers serve, an analysis 

of several hundred enquiries and 

projects has revealed correspondences 

between the detailed requirements. 

On this basis EDM has developed 

a modular platform concept 

adaptable to particular customer needs. 


exist alongside the x86 standards? Page 30 

This article reviews the proliferation of 

standards following x86, so that it is 

now difficult to define a standard in the 

clear x86 market and even more so in 

the non-x86 market, and argues that 

proprietary systems have a role in the 

non-x86 market, providing the full 

functionality of the chosen processor on one hand and the 

interchangeability of x86 systems on the other. 


through unit testing Page 35 

Unit test tools have long provided commercial benefit for the team 

developing the highest integrity applications. Now these tools can 

also streamline the efforts of their peers working in less critical 

environments – even those charged with the ongoing development 

of undocumented legacy code.


and aerospace applications 

By James B. Doyle, Emerson Network Power 

This article explains why ACTA 

is an ideal platform for system 

developers seeking rapid 

development at low cost. 

Further, the flexibility of 

the architecture gives ATCA 

the ability to meet many 

different military networking 

needs, helping ensure interoperability 

among diverse 

systems for information 

exchange and avoiding 

stovepipe installations. 

n A significant change in the military equipment 

market is a shift in focus for equipment 

development. Electronic warfare driven by 

C4ISR is gaining favour over big-ticket weapons 

systems. Winning unconventional conflicts requires 

that both the warfighter and upper 

command have rapid access to actionable intelligence. 

And that requires high-performance, 

data networking. Ideally, this data networking 

would operate across the entire force structure 

so that all service branches can exchange C4ISR 

information. Further, the networking should 

operate across multi-national forces. This need 

for wide interoperability calls for networking 

solutions that stem from open standards rather 

than proprietary designs. The ideal solution 

would be a single architecture that could serve 

a wide variety of nodes across the services and 

be able to evolve as application needs change. 

These COTS infrastructures do not necessarily 

need to handle the extreme conditions of the 

battlefield, however. Shipboard computing centers, 

airborne communications platforms, and 

theatre command centers are much more benign 

environments than field equipment must 

handle, and equipment for these installations 

do not need the same extreme levels of ruggedization. 

The network operations center of a 

naval vessel, for instance, resides on an isolated 

deck deep within the ship in an environment 

similar to commercial IT installations. While 

this environment is more demanding than 

faced by typical consumer electronics, it is not 

as harsh as that faced by UAVs, tanks, Humvees 

and the like. While such installations can utilize 

equipment with less than full military ruggedization 

they still demand that the equipment 

provide highly reliable operation. To achieve 

this, the COTS designs must incorporate high 

availability features such as fault tolerance, automatic 

switchover to redundant systems during 

hard failures, and support for rapid, hotswap 

component replacement. Fortunately, the 

evolution of the AdvancedTCA or ATCA COTS 

networking design architecture from its original 

backplane bandwidth of 1 GbE to the now 

widely-deployed 10 GbE and emerging 40 

GbE performance addresses all these operational 

requirements. 

The ATCA architecture is an open industry 

standard originally developed to offer an alternative 

to proprietary designs in the telecommunications 

industry. The standards allow 

multiple vendors to develop individual system 

components that are interoperable, allowing 

construction of systems through the mix-andmatch 

of components. The approach is fieldproven, 

with numerous examples of installed 

ATCA-based systems with more than seven 

years of operating history. To meet the operating 

needs, including compliance to telecom 

Network Equipment Building System (NEBS) 

DEFENCE & AEROSPACE 

Figure 1. Application Spectrum - the ideal 

COTS solution to the military networking 

needs would be able to tie together diverse 

nodes across all service branches. 

standards, ATCA was designed from the ground 

up with a number of key features, all of which 

are applicable to the military requirements. 

The architecture is modular, allowing considerable 

functional diversity to arise from the assembly 

of relatively few component pieces. 

The architecture is also highly flexible, supporting 

a wide variety of networking standards 

over a configurable serial backplane. 

The ATCA architecture is also more robust 

than typical industrial computing structures. 

ATCA targets high-reliability installations with 

specifications that require built-in support for 

live insertion and hot swap of modules and 

boards as well as system components such as 

power supplies and cooling fans. Fault detection 

and system control are built in, allowing module-level 

isolation and replacement as well as 

support for fail-over operation when using redundant 

designs. Even the system software 

targets high reliability, with a robust real-time 

operating system and high-availability middleware 

available as a standard software base. 

ATCA specifications also call for system operation 

in harsher environments than the typical 

home or office. Because telecommunications 

equipment must often run unattended for 

months at a time in tight quarters such as 

small cinder-block buildings, the ATCA specifications 

call for operation at sustained temperatures 

as high as 55°C, with the ability to 

5 September 2010


Figure 2. ATCA includes a built-in system management function that provides the essential features 

of fault-tolerant, high reliability system design, as illustrated by the architecture of the 

Emerson Network Power Centellis 4440 40Gbit-ready ATCA system platform. 

maintain operation for several hours even in 

the event of failures in the cooling system. For 

shock and vibration, the specifications require 

that equipment survive and continue functioning 

through a major earthquake. 

A high-level look at the ATCA architecture reveals 

how it achieves these goals. The basic 

building block of an ATCA system is the blade, 

which inserts into a high-speed switched serial 

backplane. Each blade has access to as many 

as 21 serial channels, called lanes, with each 

lane handling data at rates as high as 12 

Gbit/sec. This gives a combined data throughput 

capacity of 2.4 Terabit/sec for an ATCA 

system. The switched-serial backplane is protocol-agnostic 

allowing it to support a variety 

of serial communications formats, including 

10 Gbit/sec Ethernet (10 GbE), Serial RapidIO, 

and PCI Express. An ATCA backplane can 

support virtually any serial connection to outside 

equipment in its native format and provides 

guaranteed bandwidth for inter-blade 

Figure 3. Tests by Northrop Grumman 

demonstrated that ATCA equipment is able 

to survive barge testing, showing its suitability 

for deployment as shipboard systems. 

MIL-S-901D Heavyweight Shock Test 

Photographs courtesy of National Technical 

Systems in Rustburg, VA. 

communications. The appeal of Ethernet-based 

networking in the digitized battlefield is clear. 

And with the advent of 10 GbE and beyond, 

Ethernet is no longer limited to the command 

and control fabric in systems. It can also function 

as a fat pipe that reaches all the way down 

to high-bandwidth sensors. So it is not surprising 

that 10 GbE is being designed into 

more and more defence applications, both in 

the main network and as the pipe to and from 

sensors or effectors. 10 GbE-based ATCA systems 

can also support a wide range of military, 

aerospace and government applications including 

sophisticated adaptive beam-forming 

for satellite communications, test equipment 

with multiple processing elements, long-range 

radar systems, and high-bandwidth low-latency 

wireless test beds. 

An integral part of an ATCA system is its 

built-in system management. Each ATCA blade 

or AMC module on an ATCA blade communicates 

with a shelf manager for monitoring, 

setup, and control. Other field-replaceable 

units, such as cooling units and power modules, 

also have management controllers. The management 

controllers have the ability to determine 

module type for electronic keying. They 

can also detect failures and shut down operation 

of their client to prevent a single failure 

from affecting system operation. AMC modules, 

for instance, have independent power connections 

for the MMC and the rest of the module, 

with the MMC having control over the I/O 

drivers on the AMC. This allows the MMC to 

disconnect the AMC from the system backplane 

and power it down in the event of failure or 

the need for hot-swap replacement. Similar 

capabilities are built into the cooling units 

and power modules. Higher-level software can 

use this infrastructure to implement such functions 

as fail-over to redundant hardware for 


high-availability operation. The goal behind 

the creation of ATCA was to give networking 

equipment providers a robust alternative to 

proprietary designs to both speed development 

cycles and reduce cost while maintaining the 

ability to implement product differentiation. 

Fulfilling that promise required the creation of 

a robust COTS ecosystem to provide multiple 

sources for system components while ensuring 

interoperability among the offerings. This ecosystem 

is now both substantial and well-established. 

The PCI Industrial Computer Manufacturer 

Group (PICMG) developed and maintains the 

ATCA standard standards, including definition 

of chassis, modules, boards, and system management 

operation. More than 60 companies 

belong to PICMG and provide a wide range of 

ATCA blades, AMC modules, chassis, and full 

systems. This diverse supplier base not only 

gives developers multiple sources for many system 

components, it allows developers to choose 

the best in class for each component. The Communications 

Platform Trade Association (CP- 

TA) provides definitions and test procedures to 

ensure thermal and system management interoperability 

among these ATCA products. Other 

industry organizations provide standards and 

reference implementations for operating systems 

and high-availability middleware. 

In addition to its telecommunications origin, 

ATCA has also proven itself in applications 

such as medical systems, enterprise data centers, 

financial systems, and the military. For example, 

an ATCA-based system is currently providing 

ISR support in the Multi-mission Maritime 

Aircraft (MMA) system. Systems based on 

ATCA are also under evaluation for programs 

such as CANES. Northrop Grumman recently 

performed afloat shock testing, better known 

in the defence industry as barge testing, on 

two Emerson Network Power ATCA systems 

to verify their continual operation in a shipboard 

environment throughout a blast scenario. 

The equipment showed no faults over a series 

of tests involving multiple blasts and a variety 

of equipment mounts and orientations. 

Along with being a proven architecture, ATCA 

is continually evolving to take advantage of 

new technologies as well as expand its application 

range. The modular nature makes upgrades 

of processors and adding new interfaces 

a relatively simple matter. The architecture 

itself is also seeing upgrades through the continual 

efforts of the PICMG organization, 

which has established a new subcommittee to 

develop the standard – named ATCA Extensions. 

The brief for ATCA Extensions is fairly 

wide-ranging, encompassing from simple 

changes to make existing blades more cost-optimized 

for non-central office environments, 

through to supporting new higher computedensity 

system configurations, including dou-

Figure 4. The current ATCA specification allows the combination of four lanes to provide a 10 

Gbit/sec link between boards, and specifications are under development to extend this to achieve 

40 Gbit/sec 

ble-wide server blades to make use of bigger 

heat sinks to further increase CPU power, and 

back-to-back enclosures to make more use of 

typical data center rack depths. However, backwards 

and forwards compatibility remains a 

key driving principle, preserving existing investment 

and the ability to make use of other 

components in the ecosystem to arrive at the 

best solution. ATCA backplane speed increases 

are also under development. The PICMG or- 

ganization has already adopted an approach 

for providing a 10 Gbit/s serial channel across 

the backplane using four 2.5 Gbps lanes operating 

together. It is now developing new specifications 

for connectors and backplanes that 

would allow the same structure to utilize 10 

Gbit/s speeds on those four lanes to create a 40 

Gbit/s channel. In anticipation of this specification, 

Emerson Network Power is now offering 

chassis systems that are 40 Gbit/s-ready to 


simplify system upgrade when corresponding 

blades and AMC modules become available. 

Such development efforts by the PICMG organization 

reflect a commitment within the 

ATCA ecosystem to continually improving the 

specification and keeping abreast of new technology 

opportunities. For military applications, 

this commitment translates to assurances that 

ATCA-based systems will be upgradeable to 

track the state-of-the-art for many years to 

come. Routine upgrades can be handled through 

module and blade replacement and, while major 

upgrades may occasionally require chassis replacement, 

blades and software should remain 

compatible across most such replacements. 

The ATCA architecture thus meets all the key 

criteria for COTS-based military system designs. 

It is designed specifically for high-availability 

applications in moderately harsh environments 

similar to many military infrastructure installations, 

and is undergoing extensions to add 

ruggedization options for even harsher environments. 

Its modularity gives it the flexibility 

to adapt to a wide variety of data protocols 

and I/O interfaces, while simplifying maintenance 

and supply by minimizing the number 

of building blocks needed across multiple 

system designs. n



avionics systems development 

By Michael J. Randazzo, DDC 

CompactPCI is a first choice 

for modernizing legacy test 

equipment or developing a 

new system. Its modular hardware 

design, wide range of 

COTS I/O options, good 

instrumentation control, and 

clear migration path to future 

technology make it a smart, 

cost-effective and flexible 

solution able to support 

avionics systems development 

for years to come. 

n The CompactPCI (cPCI) form factor has 

continued to allow avionics system designers 

to keep development costs low through the 

use of commercial off-the-shelf (COTS) products. 

It remains a very popular form factor 

choice for systems requiring multiple and 

modular I/O capabilities, specifically avionics 

protocols, such as MIL-STD-1553 and ARINC 

429. In the defense/aerospace sector, some 

sample applications are system simulation/integration, 

production/automated test (ATP), 

portable test equipment, and rapid prototyping 

of custom hardware designs. 

The cPCI specification was developed by the 

PCI Industrial Computer Manufacturers Group 

(PICMG) in the mid 1990s to incorporate 

PCI-based technology into industrial computers. 

It is electrically a superset of desktop PCI 

with a different physical form factor. CompactPCI 

leverages the Eurocard form factor, 

popularized by the VME bus, and supports 

both 3U (100mm x 160mm) and 6U (160mm 

x 233mm) card sizes. Though mainly used in 

lab environments, a conduction-cooled form 

factor variant exists to support harsh shock, 

vibration, and temperature conditions, while 

maintaining the same electrical PCI characteristics 

and signaling. Since cPCI uses a chassis 

rack-mountable backplane and is physically 

similar to VME, it was quickly adopted by the 

avionics community, which had been domi- 

nated by VME since the early 1980s. The cPCI 

form factor maintains core system requirements 

such as vertical card orientation, positive card 

retention, excellent shock and vibration 

characteristics, and front or rear I/O options. 

What really added to its appeal was the use of 

standard PCI silicon, which was already being 

manufactured in large volumes for use in the 

desktop computer marketplace. This greatly 

reduced the cost of manufacturing cPCI cards 

compared with VME cards. As VME designs 

become more difficult to manufacture and 

support, cPCI offers an ideal alternative to 

support the needs formerly handled by VME 

cards. 

Once the test and measurement community 

started to adopt cPCI, it was clear that they 

needed more capability to meet timing and 

synchronization requirements across multiple 

devices. Previously, VME cards addressed this 

need with the VXI Bus enhancement. This resulted 

in the creation of the PXI (PCI eXtensions 

for Instrumentation) platform. The PXI 

platform added an additional optional connector 

to the already defined cPCI form factor. 

This connector enabled integrated timing and 

synchronization that is used to route synchronization 

clocks and triggers internally through 

the chassis backplane. Today, most test applications 

that previously would have required 

VME cards now rely on lower cost cPCI cards. 


Figure 1. Example of a 

CompactPCI chassis 

Data Device Corporation (DDC) provides 

CompactPCI/PXI cards for MIL-STD- 

1553/1760, ARINC 429, and synchro/resolver 

with many cards combining different types of 

I/O on one board to save even more power, 

space, weight, and cost. For example, the latest 

generation of Multi-IO CompactPCI/PXI cards 

BU-67107T would require only a single 3U 

card to support 4 dual-redundant MIL-STD- 

1553 channels, 8 ARINC-429 receivers, 4 

ARINC-429 transmitters, 6 user-programmable 

digital discretes, 2 RS-232 channels, and 2 RS- 

422/485 channels. Using only single-IO solutions, 

an equivalent system would require 4 

separate devices, greatly increasing the cost 

and installation time expended by system 

designers. 

With respect to the MIL-STD-1553 protocol, 

DDC gives more control to the system designer, 

by offering two levels of functionality. Its latest 

1553 generation, AceXtreme, now comes in 

single and multi-function configurations. The 

single-function configuration offers emulation 

of a bus controller (BC) or up to 31 remote 

terminals (RTs). Either mode can concurrently 

execute the bus monitor (MT), to capture all 

data on the bus. The multi-function configuration, 

targeted for test and simulation systems, 

adds the ability to concurrently run the BC, 

up to 31 RTs, and MT, and also includes the 

1553 test and simulation toolkit. The toolkit

Figure 2. 6U (Synchro/Resolver) vs. 3U (MIL-STD-1553) CompactPCI cards 

Figure 3. DDC Multi-IO MIL-STD-1553/ARINC-429 3U cPCI/PXI card 

adds error injection, internal/external triggering, 

and the ability to test out-of-bounds 1553 

parameters (such as intermessage gap). The 

BU-67210T 1553 multi-function series are 3U 

CompactPCI/PXI cards with up to 4 dualredundant 

1553 channels, 8 digital and 8 avionics-level 

discretes, and IRIG-B input/output. 

Since it is relatively easy to assemble and procure 

a CompactPCI system with COTS hardware, 

it is a great choice for prototyping custom 

embedded systems. This allows system designers 

to configure a system with identical hardware 

and software functionality as the final 

system, except using the CompactPCI form 

factor. It also enables any software development 

to begin before initial prototype hardware is 

available so that the designer can address any 

system level concerns before mass production. 

All DDC software development kits (SDKs) 

for MIL-STD-1553 and ARINC-429 are hardware 

independent and have a common API. 

This enables software to be developed and 

tested using one form factor (such as cPCI), 


and reused in its entirety on a different form 

factor (such as PMC or component-level 

solution). Designers using this approach have 

benefited from shortened development schedules, 

less system troubleshooting, and avoiding 

costly board re-spins. 

To simplify the software development process 

even further, DDC offers code generation 

capability with its BusTrACEr BU-69066S0 

MIL-STD-1553 data bus analyzer. The tool allows 

the user to configure a 1553 bus controller, 

multiple remote terminals, and a bus monitor 

with an easy-to-use point-and-click graphical 

interface. Once the 1553 configuration is complete, 

a single click will generate ANSI C source 

code to duplicate the same configuration using 

DDC 1553 SDK. The generated source code 

then can be used with any operating system 

(Windows 2000/XP/Vista/7, Linux, VxWorks, 

etc.) and any DDC board or component, saving 

the designer many hours of additional programming 

time for the embedded system that 

would otherwise be required. 

9 September 2010


Figure 4. Hybrid cPCI/cPCI Express backplane 

All CompactPCI/PXI systems require a CPU 

card or system controller. Following the COTS 

approach, there are many CPU vendors in the 

marketplace offering standard 3U and 6U controllers. 

A majority of these systems are x86based, 

but other processor families, like Power- 

PC, can be procured. Since the controllers are 

based on standard processor architectures, cPCI 

systems designers have a wealth of choices regarding 

operating systems and programming 

environments. One environment that stands 

out as a leader in the cPCI/PXI community is 

n Curtiss-Wright adds network fabric 

analysis tool to debug suite 

Curtiss-Wright adds the Network Fabric Analysis 

(NFAT) tool to its Continuum Insights 4.1 

suite of GUI-based system monitoring, event 

analysis, and system management and multiprocessor 

debug software tools. The NFAT directly 

addresses the needs of Serial RapidIO 

system architecture designers. 

News ID 10777 

n Kontron: COTS backplane offers 32 PCIe 

lanes and three PCI slots 

The new Kontron 4U backplane xPB-13E9P3 

for PICMG 1.3-based system designs boasts 

comprehensive PCI and PCI Express Gen 2 

the LabVIEW programming language. Short 

for laboratory virtual instrumentation engineering 

workbench, LabVIEW was developed 

by National Instruments in the 1980s as a 

visual programming language. Execution is determined 

by the use of a graphical block diagram 

on which the programmer connects different 

functions by drawing wires. The inherent 

parallel execution and multi-processing makes 

it a great fit for data acquisition, instrument 

control and test automation. Combined with 

extensive support for COTS hardware, a test 

Product News 

support for full-height 4U systems. With double 

the PCIe lanes of any previous solution, 32 in 

total, the PICMG 1.3-compliant backplane 

tackles bandwidth-intensive applications with 

its cost-effective and densely-packed design 

for use with standard PCIe and PCI extension 

cards. 

News ID 10795 

n MEN: 3U CompactPCI Ethernet diagnosis 

buffer 

Diagnosis buffers are used for avoiding errors 

and discrepancies in the system in nearly all 

demanding applications. MEN Mikro Elektronik 

now enters the market with its first 3U 

CompactPCI Ethernet diagnosis buffer. The 

ore information about each news is available on 

Mwww.Embedded-Control-Europe.com/bs_magazine You just have to type in the “News ID”. — 


system can be developed rather easily, without 

sacrificing capability. DDC LabVIEW support 

package BU-69093S0 enables developers to use 

LabVIEW to develop applications using MIL- 

STD-1553 and/or ARINC-429 protocols. To 

make development even easier, a collection of 

“Intermediate Vis” was created, which allow 

the user to quickly configure commonly used 

1553/429 functionality. Combined with a DDC 

cPCI/PXI hardware device, the DDC LabVIEW 

package is a good choice for any avionics 

testing or development task. The DDC package 

also supports LabVIEW Real-Time (RT), which 

extends the LabVIEW framework to deliver 

deterministic, hard real-time performance by 

executing all time-sensitive application code 

on an embedded target running the Interval - 

Zero (Phar Lap) ETS real-time operating system. 

With the advent of PCI Express (PCI-E) as the 

replacement for the PCI bus in desktop computing, 

CompactPCI needed to evolve accordingly. 

CompactPCI Express was released by 

PICMG in 2005, keeping the Eurocard form 

factor, but replacing the parallel PCI Bus with 

serial PCI-E links. Due to the wide usage of 

CompactPCI systems and peripherals in the 

field, PICMG defined the CompactPCI Express 

backplane and connectors not only to support 

new cPCI Express cards, but also support 

existing cPCI cards. A hybrid peripheral slot 

can support either a Type 2 cPCI Express 

board or a 32-bit cPCI board, thereby enabling 

any cPCI designer to have a cost-effective 

evolutionary path; cPCI cards can be eventually 

replaced with cPCI Express cards without 

changing the backplane or cabling. n 

F218 is a slave CPU board in 3U CompactPCI 

format and was developed for use as an Ethernet 

diagnosis buffer. 

News ID 10866 

n ADLINK: PICMG 1.3 SHB supports 

Core i7/i5/i3 processors 

ADLINK presents the NuPRO-E330, the latest 

PICMG 1.3 full-sized System Host Board 

(SHB) in its product lineup. The NuPRO- 

E330 supports next generation Intel Core 

i7/i5/i3 processors at clocks speeds up to 3.33 

GHz and dual-channel DDR3 1066/1333 MHz 

memory up to a maximum of 8GB in two 

DIMM slots. 

News ID 10986


Ruggedized VPX provides extra-robust 

systems for demanding applications 

By Paul Rutherford, Schroff/Pentair 

This article describes the new 

series of ruggedized VPX 

systems developed by Schroff, 

featuring a modular case with 

a suitable backplane, an 

upgradable cooling solution, 

power supply and other 

accessories such as clamshells 

- conduction-cooled assemblies, 

and enabling 3- or 6-U 

height sub-assemblies. Figure 1. Modular ruggedized 

VPX systems for extreme 

requirements 

n The VPX standard from VITA is a further 

development of the VMEbus. VPX was primarily 

created to secure the advantages of highspeed 

serial architectures for VME systems 

and thus to improve their performance capabilities. 

Particularly in aviation and space travel, 

in military applications and lately in railway 

systems, there is interest in the VPX standard. 

For the often extreme operating conditions 

and requirements in such situations as regards 

shock and vibration resistance and cooling, 

Schroff has developed ruggedized VPX systems 

that meet VITA 48.2. The VITA 48.2 standard 

describes the requirements for a conductioncooled 

version of a chassis. For Schroff this 

was the starting point for developing the new 

series of ruggedized VPX systems. These consist 

of a modular case with a suitable backplane, 

an upgradable cooling solution, power supply 

and other accessories such as clamshells (conduction-cooled 

assemblies) and Wedge- 

Loks/Card-Loks. The modular nature of the 

new system is in keeping with the renowned 

Schroff platform concept and allows users not 

only to build systems on a modular basis but 

also allows subsequent expansion and/or upgrading 

of the systems. The systems are also 

flexible in terms of dimensions, the cooling 

solution and the EMC and IP protection. The 

case of the ruggedized VPX systems is formed 

of machined aluminium parts (base, top cover, 

side, front and rear elements) bolted together. 

Various aluminium alloys can be used (EN 

AW6062, 6082, 7075, etc) as specified by the 

customer. A variety of finishes are also available: 

black anodised, nickel-plated, yellow chromated 

etc. A grid in the side components, top cover 

and base allow other optional parts such as 

mounting brackets to be attached. The front 

and rear elements are symmetrically designed 

and have identical hole positions and bolt-on 

dimensions. Custom cut-outs can be integrated 

here for specific requirements (for connectors, 

switches etc). The area available for this extends, 

thanks to a very narrow fixing flange, over almost 

the entire internal dimensions of the 

case. The front and rear cover elements are 

available in various depths. This allows the installation 

space for the boards to be expanded 

towards the front or rear. In its basic configuration 

the case contains no gasketing. Where 

requirements dictate, however, either a pure 

IP seal (to protect against the ingress of moisture, 

dust etc) or a combined IP and EMC seal 

can be fitted, ex-works or as a retrofit, in a 

groove provided for this purpose. 

The VPX systems can be prepared for sub-assemblies 

of 3 or 6 U heights. The board format 

is both the standard and the double Euroboard 

format (100x160mm and 233x160mm). Customers 

who previously used VME or Compact- 

PCI systems can, in converting to the VPX standard, 

fall back on the familiar form factor. The 

guides for the PCBs are milled into the base 

and top cover of the case. Here VPX defines 

three pitch increments - 0.8“, 0.85“ or 1“ - for 

the separation between boards in a system. 

Schroff offers all three options. The physical 

construction keeps the thermal interfaces in 

the case to a minimum. This is because each interface 

transfer from one component to another 

increases the thermal resistance and is thereby 

detrimental to the effective dissipation of heat. 

Cooling for these new ruggedized VPX systems 

can be upgraded through four levels. The first 

level is the simple basic case as already described. 

Components housed in the case are 

cooled by convection alone. Heat is drawn 

from the PCB via a simple board frame or a 

closed frame (clamshell) to the surface of the 

case. In the second level, heat sinks are fitted 

to one or more sides of the case exterior. 

Thanks to the ribs of the heat sinks, the overall 

surface area of the system is increased and 

thus also the heat dissipation from it. In level 

three a sheet metal cover is placed over the 

heat sinks and a fan with suction chamber is 

fitted at the rear of the case. The presence of 

the cover creates an air channel. The air is 

sucked by the fan through the heat sinks and 

drawn to the rear. This forced air cooling further 

increases the heat dissipation capacity of 

the system. For the fourth and highest level the 

heat sinks are replaced by liquid-cooled plates. 

The use of a liquid cooling medium multiplies 

the cooling capacity considerably. The cooling 

liquid draws the heat from the case surface 

and is pumped to, for example, a chiller unit. 

All cooling elements involved in the upgrade 

11 September 2010


Figure 2. The metal cover provides an air 

channel; the fan draws the air through the 

heat sinks and away to the rear. 

Figure 3. Two clamshell versions: primary 

side (shown on right) and secondary side 

(shown on left) 

levels - heat sinks, cover and liquid-cooled 

plates - are fixed to the case using the same 

fixing points. The modular cooling concept 

thus enables the user to adapt the VPX system 

to the current cooling requirement at any time. 

In the design and development phases in particular 

the precise cooling requirement may 

not yet be known. With this modular design, 

however, the cooling capacity can be upgraded 

or adapted as required without problems. 

Conduction-cooled assemblies (CCAs), Wedge- 

Loks/Card-Loks and clamshells have also been 

developed for particularly demanding applications. 

These are used on the one hand for securing 

PCBs in harsh environments, e.g. on 

ships, aircraft and trains, in mobile communications, 

etc, and on the other hand serve an 

important function in drawing away the heat 

generated on the boards. CCAs are board 

frames in aluminium that are joined to the 

boards. However, heat transfer by this means 

to the case surface is only achieved at localised 

points. Clamshells, described in VITA 48.2, are 

fitted top and bottom with Wedge-Loks/Card- 

Loks and an insertion/extractor handle; in 

addition to fixing the boards securely into the 

system, they offer a far more effective means of 

heat dissipation through their large surface 

contact. As a rule they encase plug-in unit 

PCBs from both sides. Two versions are common: 

with a smooth inner contour and with 

an inner contour milled to customer specification. 

The milled contour forms a negative 

impression of the outline of the components 

on the board. Contact points are thus formed 

between the internal surface of the shell and 

the heat-producing components. For optimal 

heat transfer, special elastic, heat-conductive 

pads are additionally employed. Thus the heat 

passes directly to the clamshell and then via its 

flange to the case surface and outwards. 

VITA 48.2 defines two versions of clamshells 

in terms of their exterior contour: the primary 

and the secondary sides. This describes the 

side on which the Wedge-Loks/Card-Loks are 

mounted or that on which the components 

with the greater heat production are located 

(primary = right, secondary = left). The side 

with greater heat generation is always without 

Wedge-Loks/Card-Loks so that fewer interfaces 

are encountered on the way to the case surface, 

thus keeping thermal resistance to the minimum 

possible. Clamshells are accordingly available 

in widths of 0.8“, 0.85“ and 1“ and in 3 and 6 

U to correspond to the guides in the case. 

The clamshells can optionally include extra 

protection for the connector. This is beneficial 

if, for example, two-level maintenance is required, 

as is the case in the US military. Here 

different maintenance levels and their various 

categories of protection are defined. At the 

highest level it must be possible, in the event 

of failure, to replace PCBs in the field, under 

difficult conditions, without the danger of the 

operator damaging anything. For this special 

Wedge-Loks/Card-Loks are required, such as 

are provided by Schroff’s sister company Calmark. 

A new series features torque limiting 

for which a patent has been registered. Now 

boards can be fitted into the chassis with a 


prescribed clamping force, and without the 

need for a torque wrench. The new Card-Loks 

are designed such that no damage can be done 

to the board or to the Card-Lok itself by excessive 

tightening. When the screw is tightened, 

an audible click indicates when the predetermined 

torque has been reached. If the operator 

tries to tighten further, the screw slips, by 

means of a ratchet clutch; it is not possible to 

overtighten the screw, or to break it off. 

The ruggedized systems are currently fitted 

with a 5- or 7-slot full-mesh backplane. Other 

numbers of slots and topologies in line with 

the Open VPX specification are in preparation. 

The VPX backplanes are, like other high-speed 

Schroff backplanes (AdvancedTCA, MicroTCA 

and CompactPCI) designed for 10 Gbit/s on a 

differential pair or 40 Gbit/s on four differential 

pairs. Verification of these transmission rates 

has been made in in-house laboratories in the 

time domain (TDR) and the frequency domains 

up to 20 GHz, and with a 10 G pattern generator. 

There are VPX backplanes for ruggedized systems 

and for conventional 19“ subracks e.g. 

for development systems for laboratory testing. 

In both cases the initial product is one PCB 

only. Adaptation of the backplane from the 19“ 

subrack form factor to the ruggedized form 

factor is simple. In the 5- and 7-slot backplane 

two slots are provided for power supply units. 

Since there is currently still no plug-in PSU for 

VPX on the market, the backplanes are 

equipped with powerbugs, allowing the use of 

existing open-frame power supply units. As 

soon as VPX PSUs become available the 

backplanes will be modified accordingly. n 

VME is still the dominant architecture in the embedded COTS boards market and will remain 

so for a very long time. But ICC Media’s free virtual conference on September 28 will not only 

cover the traditional VMEbus market, but will focus on advanced technologies and “hot” topics. 

VME Trends & technology Updates 

n Business conditions & technical trends with 

critical embedded systems 

Interview with Ray Alderman, Executive 

Director VITA 

n VME Technologies - a status-update and technical trends 

Speaker: Jerry Gipper, VITA 

n VME market update 

Speaker: Richard Dean, Program Manager VDC Research 

n VPX Technology - Technical Trends 

Speaker: Neil Peterson, Chairman of the VPX Marketing Alliance 

n Flexible I/O in a rigid world - an introduction to the FMC (VITA 57) standard 

Speaker: Malachy Devlin, VITA 57 Chairperson 

n VXS Switched Serial Architecture 

Speaker: Justin Moll, Chairman of VXS Marketing Alliance 

Applications, Products & Solutions 

n Application- and product-oriented presentations from the sponsors of the conference 

www.iccmedia-vcon.com


in the trains of the future? 

By Manfred Schmitz, MEN 

This article discusses 

the various train busses 

used today and the 

possibility that Ethernet 

could become the train 

bus of the future. 

n Railway vehicles – be it railcars or locomotives, 

underground trains, trams or passenger 

and freight trains – are equipped with a multitude 

of single functions which are all interconnected. 

For the train drive these are the 

control of the traction and the brakes, the 

clutches and the gearboxes, the antiskid system 

and the increase of braking power via sanding. 

In the driver cabin there are the display systems 

for speed, tank filling level, SIFA (monitoring), 

GSM as well as the data logger and train safety 

systems. The motor starter, the pre-heating 

system or the traction cooling system are connected 

to the main supply, while the electric 

equipment and the battery are connected to 

the auxiliary supply. 

In a vehicle for passenger traffic there is also 

the control of the inner and outer doors, the 

inner and outer lighting system, the air conditioning 

and the sanitary equipment, the train 

coupling and possibly the tilting system. Recently 

passenger information systems and wireless 

access to internet services have been increasing 

traveling comfort. Another trend is 

to systematically equip trains with monitoring 

and safety systems for passengers and the train 

infrastructure. In the past, many of these subsystems 

could be accessed via point-to-pointconnections 

on the basis of RS232 or RS485. 

As the number of computer-controlled subsystems 

increased, it was necessary to switch 

to higher-performance communication channels 

so that different fieldbus networks were 

introduced in railway vehicles. The latest dataintensive 

innovations like infotainment, web 

access and other wireless services like GPS 

and GSM led to solutions based on Ethernet 

networks. At the moment the whole transport 

industry is changing rapidly, there are as many 

approaches as there are vehicle manufacturers. 

In the IEC TC 9 WG 22, the train communication 

network TCN was created in the 90s as a 

railway-specific fieldbus according to IEC 

61375, in cooperation with the UIC (International 

Union of Railways). It consists of the 

vehicle bus MVB (multifunction vehicle bus) 

and the train bus WTB (wire train bus). The 

TCN defines the time and safety critical data 

transfer in railway vehicles using the complete 

layer 7 architecture. Within the TCN network 

the MVB can be used as a fieldbus inside one 

vehicle or inside a group of fixedly coupled vehicles. 

It is defined for a data transfer rate of 

1.5 Mbits/s via optic fiber cables and/or twisted 

pair cables. As the MVB controller chips are 

proprietary and expensive, often other fieldbuses 

like CANopen or Profibus are used instead 

of the MVB. The WTB train bus, a dynamic 

bus system that was developed for operation 

of passenger trains with a locomotive, 

is also extremely expensive in the application. 

It features exceptional processes such as the 

TRANSPORTATION 

Figure 1. Rugged Ethernet 

switch for use in trains 

automatic numbering of the wagons during 

train formation or the building up of fritting 

voltages (very high direct voltage used for 

cleaning of coupling contacts). It is defined 

for a data transfer rate of 1 Mbit/s over a distance 

of up to 860 meters via a twisted pair 

cable and built up redundantly. 

Seen from today, the bandwidth of TCN and 

the other fieldbuses is reaching its limit. There 

are no user organizations as for CANopen 

and Profinet which look after the standard. 

Thus there are implementation differences 

and differing incompatible physical interfaces 

which complicate interoperability. In addition, 

the cost per node is ten to one hundred times 

higher than for CANopen. This entails that 

dependent on the manufacturer different fieldbuses 

like CANopen, Profibus, Profinet, World- 

FIP, Bitbus etc are used. This in turn generates 

unnecessary costs, as different fieldbuses require 

different configuration tools and analyzing 

devices. In more complex trains, this 

requires gateways which have to mediate between 

the different buses and protocols. Price 

and bandwidth problems together make it 

ever more necessary to use high-performance 

standardized transport protocols like TCP/IP 

or UDP/IP. The big technological progress of 

Fast, Gigabit and 10Gigabit Ethernet, switching 

and collision-free full-duplex transmission as 

well as time-synchronous protocols and stan- 

13 September 2010

TRANSPORTATION 

Figure 2. Different train applications today linked by various networks 

dardized safety protocols on Industrial Ethernet 

has accelerated development in automation, 

which makes it possible to replace the fieldbuses 

used today. In real-time applications, 

however, there are competing standards like 

Profinet, EtherNet/IP, EtherCAT, Ethernet Powerlink 

and many more. Which solution will 

establish itself in railway vehicles remains to 

be seen, but the costs per node should be similar 

to those of CANopen. This opens up new 

possibilities to build a completely homogeneous 

network based on Industrial Ethernet 

onboard a vehicle. 

Figure 3. Rugged connectors for use in train 

applications 

Which (real-time) Ethernet will come out on 

top in trains mainly depends on it being 

capable of meeting some very special requirements 

regarding functionality, environment 

and maintenance. In trains three network 

topologies are used; line, ring and ladder structures. 

While line structures are chosen for uncritical 

data, rings are common in real-time 

environments. Gateways are indispensable as 

long as different communication buses are 

used in the train. An important function is 

fritting, i.e. cleaning the coupling contacts via 

current surges. As a simple and cost-effective 

solution, Ethernet offers the possibility to use 

Power over Ethernet. Here, the data lines are 

used for supplying end devices and overlaid 

with a direct voltage. There is an alternative for 

building up train buses in existing trains with 

couplings without data lines. It is called Ethernet 

Powerline (also DLAN or Homeplug). This 

„network via power socket“ establishes itself 

increasingly beside the traditional Ethernet 

cable and the WLAN network. One of the 

undisputed advantages is its easy installation. 

The theoretical speed of the newer Homeplug 

AV standard is 200 Mbits/s by now, that means 

it is faster than WLAN and the usual 100-Mbit 

LAN for a maximum line length of 200 meters. 

As the encoded data stream ends at the electricity 

meter, Homeplug is safer than WLAN. 

Drawbacks of Homeplug are the price, the intercompatibility 

of the devices, the electromagnetic 

susceptibility and the temperature range. 

In addition environmental requirements as 

prescribed by EN 50155 have to be taken into 

account. A robust connector technology has to 

be used on the hardware side. The RJ45 connectors 

often used for Ethernet are not accepted 

in the railway market. Instead, D-Sub connectors 

and M12 connectors which are now also 

available for Gigabit Ethernet are used. The 

power supply units have to meet special requirements, 

e.g. they have to support wide 

input ranges from 9V to 154V. The requirements 

for protective circuits also differ from those for 

standard industry components. For example, 

transients of 1800V for duration of 50 µs are 

specified. Also, different demands are made on 

the maintainability of the Ethernet devices. 

These are for example used as an easily exchangeable 

19“ cassette or as compact devices 

for space-saving wall-mounting. For an easy 

on-site configuration dongles are usually used. 

This scenario in which there are still many 

unanswered questions from perspective today 

will probably become reality much faster than 

expected. The driving factor is the big financial 

gain for the final integrators and end customers 

FREE Subscription to boards & solutions magazine 


or users. They will finally be able to use standardized 

protocols as well as a wide range of 

OEM devices and assemblies (including the 

Ethernet backbone with switches, cables and 

connectors). Product costs and system costs 

can be lowered considerably. They can benefit 

from the dynamic development in IT – while 

conforming to the industry standard EN50155. 

In the last instance the vehicle suppliers benefit 

from the basic developments and use TCP/IP 

or UDP/IP as transport protocol. They do not 

have to care for railway specific proprietary 

developments. The same scenario applies to 

the infrastructure in buses or utility vehicles. 

For operation of passenger trains with a locomotive, 

the problem of the automatic numbering 

of the wagons during train formation, 

i.e. the need for interoperability, remains, so 

the WTB will probably be used there for a 

long time still. In applications where the UIC 

556 does not have to be taken into account, 

like for the modern ICE or TGV/AGV in 

traffic, the vision of an Ethernet-only communication 

structure could become reality in the 

next three to five years. 

Ethernet as a backbone for infotainment applications 

(announcements, advertising, films 

on the one hand – mobile office, games etc on 

the other hand) in the train is state-of-the-art 

even today. Likewise, monitoring systems for 

the safety of passengers as well as logistic functions 

communicate via Ethernet. In one of 

several configuration possibilities for infotainment 

systems, communication inside the train 

is carried out via several Fast Ethernet channels, 

while wireless interfaces are used for external 

communication – for example for satellitebased 

internet access. Access to GPS, UMTS, 

GSM or HSDPA is ensured via PCI Express 

MiniCards on suitable carrier boards. All these 

functions are typically integrated modularly 

and maintenance-friendly in CompactPCI systems 

for 3U boards. For controlling the system, 

one or several Intel-based CPU boards are 

used which can carry out different functions. 

RAID configurations for content servers are 

housed in similar systems. Camera surveillance 

systems are also often equipped with one or 

several hard disk slots for building a RAID 

configuration. One or several railway-compliant 

wide-range PSUs per system complete each 

solution. Often the assemblies are coated 

against dust and humidity according to EN 

50155 and specified for an operating temperature 

of -40 to +85°C. n 

Ensure getting your personal copy of bas magazine free of charge by completting the online form at: 

• www.embedded-control-europe.com/bas-magazine 

www www.embedded-control-europe.com/ba 

p / as-magazine g

Embedded Forum 

Preview Embedded Forum at electronica - Munich, Nov 9-12, Hall A6 

The Embedded Forum is a theater-style presentation area with free access for 

all electronica visitors 

At the Embedded Forum ICC Media is staging a 4-day forum programme with technical 

presentations covering the whole range of embedded technologies from Chips & Components 

over Tools & Software to Boards & Modules. 

The Embedded Forum is located in Hall A6, the dedicated “Embedded Hall”, close to 

the Main Entrance East 

Programme Overview 

Tuesday, Nov 9th 

Session 1 - Real-Time Operating Systems 

Session 2 - Embedded Connectivity 

Thursday, Nov 11th 

Session 1 - Model-based/Graphical Design 

Session 2 - Development Tools 

Session 3 - Microcontrollers & Pocessors, DSPs 

More Information: 

Wednesday, Nov 10th 

Session 1 - Small Form Factor Boards 

Session 2 - Motor Control 

Session 3 - Virtualization & Hypervisors 

Friday, Nov 12th 

Session 1 - Electronic Lighting 

Session 2 - Software Development 

electronica.embedded-know-how.com

INDUSTRIAL AUTOMATION 

Industrial imaging trends lead 

to FPGA-based framegrabbers and GPUs 

By Reinhard Borst, Eltec 

This article highlights the state 

of the art in industrial 

imaging, and shows how 

FPGA-based solutions offer 

increasing advantages by 

reference to the Eltec range 

of image-processing 

components, including bus 

interfaces and camera 

connection technology. 

n Based on its performance needs, imaging is 

an early adopter of new technologies in the 

automation industry. Currently innovations 

are driven by the use of high-density and flexible 

FPGAs to provide the requested computing 

power, while on the other hand graphic processing 

units (GPUs) are discussed as alternatives. 

On the interface side of industrial image 

processing, Gigabit Ethernet for machine vision 

(GigE Vision) standardizes the connection of 

cameras to the cabling structure of the GB- 

Ethernet, while the Camera Link interface was 

defined by a group of industrial camera and 

frame grabber manufacturers for use in demanding 

vision applications. While interface 

technologies like USB 2.0 (480 Mbit/s, maximum 

4.5 m reach), Firewire (4.5 m reach) or 

WLAN (slow, 802.11n provides only 8 MB/s) 

show some weaknesses on performance and/or 

reach for imaging applications, there are clear 

advantages for Camera Link or Gigabit Ethernet 

(GigE Vision). 

In the past, the industrial digital video market 

has lacked a standard method of communication 

between cameras and frame grabbers. 

Camera manufacturers and frame grabber 

manufacturers developed products with different 

connectors which made cable integration 

expensive and sometimes confusing for consumers. 

Also, as data rates and the complexity 

of data transmissions continue to increase, the 

need for a connectivity standard has become 

critical. In an era of fast, complex data exchange, 

hand-built cables will no longer provide the reliability 

needed. By specifying a standard pin 

arrangement and a cable assembly that is specifically 

designed for reliability at high data rates, 

the Camera Link standard ensures that compatible 

devices can be connected with ease. 

Camera Link is a communications interface 

developed for use in vision applications. It is 

based on the Channel Link technology developed 

by National Semiconductor. Channel 

Link is the latest advance in LVDS (low voltage 

differential signaling) technology for transmitting 

digital data. Channel Link uses a parallelto-serial 

transmitter and a serial-to parallel-receiver 

to transmit data at rates up to 2.38 Gbps. 

The maximum serial throughput is 595 Mbps. 

Since it uses low-swing differential current 

mode drivers, Channel Link reduces EMI. But 

the major advantage is that multiplexing of 

wwww.embedded-control-europe-com w w. 

embedded 

-con 

n trol-eur 

ope-com 


Figure 1. PCI Express 

framegrabber PC_EYE/CL 

uses the full Camera Link 

potential 

the data lines provides a substantial reduction 

in cabling complexity. It also means that smaller 

connectors can be used on the cables. On the 

other hand the maximum reach is only 2.5m 

and a separate PC interface is needed. 

GigE Vision uses the IP transmission protocol, 

which allows GigE products to communicate 

with a computer. However, the product does 

not utilize the entire TCP/IP stack with error 

correction and flow control, as this causes 

high CPU load (approximately 1 GHz, needed 

especially in the camera). Filter drivers lower 

the CPU load to < 10% on the receiver side 

but do not run everywhere. Instead, the solution 

uses a simplified correction system and 

does not utilize the throughput control - each 

camera can only transmit at a fixed rate. 

Among the most significant advantages of 

GigE Vision are as follows. The hardware is interchangeable 

and the approach is industrial. 

Lower priced camera components can be used 

for image processing in combination with the 

familiar Ethernet cabling components. The 

Gigabit Ethernet data rate (1.25 Gbps) makes 

it possible to attain data throughput rates of 

up to 100 MBytes. Given that both the Gigabit

Figure 2. Camera module with GigE Vision interface 

Ethernet and GigE Vision are general standards, 

it is possible to use cameras from a variety of 

manufacturers for streaming functions and 

industrial image processing. Having more cameras 

in one system a separate interface for 

each camera is needed. The distance of up to 

100m that can be bridged is considerably 

longer than the distance that could be covered 

to date with competing technologies. Summarized, 

Gigabit Ethernet provides the benefits 

of wide usage, inexpensive parts and long 

cables (100m), but needs special imaging drivers 

or interfaces depending on the application. 

On the other hand Camera Link provides the 

best performance at higher costs, but only 

short cables are supported and also drivers/ - 

interfaces are needed. 

Typical methods of increasing speed in image 

processing include the distribution of the computing 

tasks between multiple multi-core 

processors, specialized FPGAs or graphic 

processors (GPUs). Each of these technologies 

has its own advantages and disadvantages. 

GPUs were originally built for graphics tasks 

like shading, reflection, and hidden surface removal. 

Becoming more easy to program, they 

are also able to perform more than specific 

graphics computations. As general-purpose 

coprocessors, the speed and architectures of 

these devices makes them useful for a variety 

of other applications including image processing 

and machine vision. However, the high 

speed based on more than 1000 processing elements 

is not fully available to image processing 

tasks because of delays in the data flow, from 

image capture to data processing. 

Algorithm design is a key criterion for the 

GPU usage in imaging applications. As Open 

CL is for parallel programs for GPUs and 

CPU kernels, Nvidia Compute unified device 


architecture (CUDA) is used for detailed programming. 

CUDA provides a set of extensions 

to the C language that allows programmers to 

target portions of their source code for execution 

on the device. Nvidia provides a host, device, 

and common component runtime as 

part of the CUDA to issue and manage computations 

on the GPU without the need to 

map them to a graphics API. When programmed 

through CUDA, the GPU is viewed 

as a computing device capable of executing a 

number of threads in parallel, allowing compute-intensive 

portions of applications running 

on the host to be offloaded onto the GPU. In 

contrast, GPUs have lots of blocking options 

(memory access). GPUs are predestined for 

use with standard PCs with high computing 

demand and deliver up-to-date performance. 

On the other hand, one of the biggest disadvantages 

of GPUs is that they are powerhungry 

with power consumptions of up to 

100W. In addition the graphics board uses the 

only available PEG (x16) slot. This means that 

an expandable PC is needed. 

The image processing components of Eltec 

have been based on FPGAs for many years. 

First bus interfaces and camera connection 

technologies were implemented into FPGAs. 

The latest projects also implement the application 

- at least pre-processing. Core competences 

include, among other things, IP core-based 

bus interfaces for PCI, PCI-X and PCI Express, 

the pre-processing of images for the index 

line correction of sensor applications, the 

direct implementation of frame grabbers on 

mother boards as part of the sensor technology, 

and the development of large reference image 

memories with SO-DIMMs in the multiple 

Mbytes range. These solutions support complexities 

of up to 1 million gate equivalents, 

pixel rates of up to 100 MP/s per channel and 

17 September 2010


Figure 3. FPGA-based framegrabbers provide 

flexible imaging solutions 

n MEN: peripheral boards for CompactPCI 

PlusIO 

The new serial bus standard PICMG 2.30 

CompactPCI PlusIO has been on everyone’s 

lips for the last one and a half years. MEN 

Mikro Elektronik now brings a number of 

different peripheral boards on the market to 

complement the product range. 

News ID 11039 

n Kontron: Box PC now available for ambient 

temperatures from -15 to +60 Â°C. 

The Kontron Embedded Box PC CB 752 is 

now available for ambient temperatures from 

-15 to +60 Â°C. The electronics are equipped 

with particularly robust specified components 

and the whole system was fully tested in the 

climate chamber. 

News ID 11028 

n COMMELL: industrial-grade Pico-ITX board 

with Atom Z510P 

COMMELL unveils its latest addition to the 

family of industrial-grade Pico-ITX (measuring 

just 10 x 7.2cm) boards. The Embedded miniboard 

LP-171 is based on the Intel Atom 

processor Z510P and the Intel System Controller 

Hub US15WP. It is especially suitable 

for space constrained applications. 

News ID 10978 

n Beckhoff: Control Panel and Panel PC 

series with 24“ widescreen TFT 

Beckhoff has extended its Control Panel and 

Panel PC series with a larger 24“ display. The 

the simultaneous connection of up to 16 

cameras. If each single pixel of a large image is 

to be subjected to the same algorithm, preprocessing 

within the hardware remains the 

only option available to date. These days FPGAs 

provide the best option for this approach. The 

development is done in VHDL and ensures 

that the resulting firmware is portable and 

works seamlessly across the boundaries between 

manufacturers. The implementations available 

to date cover iconic processing (image enhancement) 

as well as the correction of errors. 

Nowadays PCs depend entirely on fast standard 

busses, while PCI Express, the first serial bus 

of its kind, is in the process of replacing its 

predecessors PCI and PCI-X. The realization 

of PCI and PCI-X in FPGAs has been possible 

for quite some time. In this discipline, Eltec 

bets on verified IP cores made by specialized 

vendors. However, FIFOs and DMA controllers 

are also still required for image processing, as 

IP alone does not suffice. The latest generation 

of FPGAs also includes serial interfaces with 

Product News 

new widescreen display, with its high resolution 

of 1920 x 1200 pixels, is ideal for complex, 

graphical user interfaces and for information 

that must be visible from a distance. These 

visually-stunning new displays can be 

equipped with an optional touchscreen. The 

Panels can also be customized for maximum 

flexibility. The robust, water- and dust-proof 

design of the aluminium Panels makes them 

suitable for application directly on machines 

or installed elsewhere on the plant floor. 

News ID 11026 

n Eurotech helps individuals with speech 

challenges communicate 

Eurotech announces multiple contracts with 

DynaVox to deliver the Catalyst LP and Catalyst 

Module embedded computers for the DynaVox 

V, Vmax and Xpress alternative augmentative 

communication devices. Last year, DynaVox 

chose the Eurotech Catalyst Module for their 

Xpress speech-generating device. This successful 

engagement allows DynaVox to leverage 

Eurotech’s previous integration efforts for a 

faster product development cycle. 

News ID 11011 

n Advantech: Q57-based industrial 

mATX motherboard 

Advantech announces the production of its 

new Intel Q57 based industrial grade mATX 

motherboard. The AIMB-580 features Intel’s 

new 2 chip solutions for power management 

savings to increase energy efficiency, PCIe Gen 

2 for GFX add-on card support, high-band- 


GHz speeds right on the chip so that they are 

instantly suitable for PCI Express. Paired with 

the proper IP cores and proprietary DMA 

logic, the functionalities required for image 

processing can now be achieved. Using this 

approach, PCI Express is available with x1, x4 

and x8 lanes, depending on the bandwidth 

desired. 

FPGAs can also be used for the efficient and 

flexible implementation of a wide range of 

different camera interfaces (analog, LVDS, 

Camera Link, GigE Vision, fiber optics, network 

based), customer-specific frame grabbers, and 

interfaces for the installation into cameras, secure 

transmission protocols and frame buffer 

graphics. Moreover, FPGAs are predestined 

for the realization of internal FIFOs used to 

decouple data streams, and of an efficient 

DMA controller solution. Eltec does not only 

use the latest FPGA technologies in its standard 

portfolio of image processing products, but 

also for customer-specific solutions in other 

product segments. n 

width DDR3 memory support, Intel vPro technology 

with AMT6.0 for remote management, 

all in a low-power design that does not sacrifice 

performance. 

News ID 11111 

n IPC2U: full size PICMG 1.0 CPU card 

for extended temperature range 

IPC2U presents the full size PICMG 1.0 CPU 

card WSB-G41A. The card can be equipped 

with the LGA775 Intel Core 2 

Extreme/Quad/Duo CPU (1333/1066/800MHz 

FSB). Celeron CPU can be also used with this 

board. Up to 4GB 1066/800MHz DDR3 memory 

supports the high performance of these 

processors. 

News ID 10838 

n DFI: QM57 Express chipset microATX 

Industrial motherboard 

DFI launches a new microATX form factor 

motherboard, CP330-NRM, incorporating the 

latest two-chip platform using a Mobile Intel 

QM57 Express chipset supporting Intel Core 

i7-620M, Intel Core i5-520M, Intel Celeron 

P4500 processors that combine intelligent 

performance with power efficiency. 

News ID 10843 

n Janz: integrated solutions for 12- and 

15-inch emVIEW systems 

Janz offers complete integrated system solutions 

for 12 inch and 15 inch emVIEW systems. 

Other sizes are available on request. The displays 

and Panel PCs from the emVIEW family

can be built into Rittal Comfort-Panel and 

Optipanel system housings. The solutions are 

also suitable for support arm and base mounting. 

Customers can therefore ideally operate 

or observe a machine using this flexible system 

solution of placing the display panel in an optimal 

position. Whether an operator is sitting 

or standing â€“ all system elements are able 

to rotate, tilt, swivel, raise and lower. 

News ID 11084 

n EKF: CompactPCI to PCI adapter 

As an industrial standard, CompactPCI is a 

modular solution for configuring industrial 

grade PC systems. Though todays market is 

offering a broad variety of CPCI I/O solutions, 

in some cases the system integrator 

needs a special function available on a standard 

PCI card only. The CA2-FUNK from 

EKF is a CompactPCI carrier board, either 

3U or 6U front panel height, equipped with 

up to three slots for accommodating standard 

PCI cards. 

News ID 10925 

n Tews: extended temperature 3U 

compactPCI module 

Tews Technologies introduces the 4 channel 

TCP467, an extended temperature and high 

density 3U CompactPCI Serial Communications 

Controller. The TCP467 is ideal for 

applications in transportation, communications, 

process control, and COTS.The new 

module provides 4 channels of high performance 

RS232/RS422/RS485 selectable 

serial connectivity. 

News ID 10896 

n Quanmax: fan-less aluminum Box PC 

with Atom D510 

Quanmax announces the QBOX-1200 fan-less 

aluminum Box PC. QBOX-1200’s exclusive 

radian fin design provides the best cooling 

performance. Additionally, with the fanless 

and compact design, QBOX-120 can avoid 

the dust caused by convection fan Interference, 

and ensure continuous operation time. 

News ID 10968 

n Trenton: PCI Express 2.0 industrial 

backplanes 

The Trenton BPC7041 and BPC7009 industrial 

backplanes are now available for use in 

military & aerospace, virtualization and video 

display systems. These compact backplanes 

maximize high-performance computing flexibility 

by supporting up to eleven PCI Express 

2.0 and PCIe 1.1 COTS option cards is an 

industrial computer. The backplanes also 

support a wide array of PICMG 1.3 single 

board computers including Trenton’s 

JXT6966 featuring the latest quad-core Intel 

Xeon processors. 

News ID 10992 

n General Micro Systems: rugged computer 

with up to 2.26 GHz Core 2Duo Penryn 

General Micro Systems has introduced a fast 

rugged computer system, Golden Eye II, based 

on the upgradable Intel Penryn processor. 

Golden Eye II with up to 2.26 GHz Core 2 

Duo processor with 6 Mbytes of L2 Cache, 

supports up to 8 Gbytes of memory and up to 

2 Tbytes of removable storage, perfect specifications 

for a data logger/recorder or a video 

capture system. 

News ID 11037 

www.embedded-control-europe.com/newsletter 

www .embedded-control-europe.com/newsletter 


n Axiomtek: 22-inch fanless and water-proof 

Atom touch panel computer 

Axiomtek announces its IFO2225-830 fanless 

& spill-proof touch panel computer with 22inch 

WSXGA+ TFT widescreen panel, supporting 

1680 x 1050 pixel resolution and blending 

Intel Atom processor N270 1.6 GHz with Intel 

945GSE+ICH7M core logic chipset with maximum 

256MB sharing graphics memory. The 

panel supports DualView and multiple I/O options 

including 4 USB 2.0 and 4 RS-232 ports. 

News ID 10987 

Embedded News 

Free E-mail Newsletter for Europe`s Embedded Engineers 

• Chips & Components 

• Tools & Software 

• Boards & Modules 

Issued every 2 weeks 

19 September 2010



concept for custom applications 

By Ingo Brussog, Siemens EDM 

Despite the varied purposes 

that industrial controllers 

serve, an analysis of 

several hundred enquiries and 

projects has revealed 

correspondences between the 

detailed requirements. On this 

basis EDM has developed a 

modular platform concept 

adaptable to particular 

customer needs. 

n Industrial controllers can be defined in many 

ways. For example, an industrial oven or heat 

pump has to be controlled. A single controller 

can perform a wide variety of tasks in a large 

number of industries. There are, for example, 

simple control tasks, which can be performed 

by any standard controller in the catalog ranges 

of countless suppliers. However, specific and 

innovative requirements demand special solutions. 

This is where Siemens EDM comes in: it 

focuses on the needs of industry. Solutions tailored 

to industrial needs are required in quantities 

ranging from a few hundred to 100,000 

per annum. Despite this high number, an industry-wide 

analysis of several hundred enquiries 

and projects has revealed correspondences 

between the detailed requirements. As 

a result of this examination, EDM has developed 

a platform and modular concept that can be 

adapted to particular customer requirements. 

A modern industrial controller generally consists 

of a main module with complex interfaces, 

the HMI (human machine interface), the associated 

high-performance processor, a suitable 

operating system, and one or more sensor-actuator 

modules. The assembly varies in individual 

cases, but some modules recur repeatedly, 

for example digital input, digital output, analog 

input as a measuring circuit, and analog output 

as a control signal. Identical software modules 

are also required, for example: remote service, 

update algorithms, data logging, and access 

rights. EDM refers to them collectively as 

system services. 

Many years of service experience with a large 

number of processors has shown that the currently 

available representatives of the ARM 

family are particularly suitable, for example 

the ARM7 and, in future, the Cortex M derivatives 

as well. They are already used as a proven 

CAD macro for the real-time handling of customer-specific 

sensor-actuator modules. The 

ARM9 platform already covers more than one 

pure CAD macro. It is available as a prefabricated 

module, complete with memory, onboard 

power supply and all current, conventional 

standard interfaces. Embedded LINUX or 

Windows CE can be offered as the operating 

system. The ARM9 platform can be used for 

connecting the widest range of displays, with 

up to VGA resolution. It is predestined for employing 

resistive touches as the means of input. 

EDM offers an ARM11 platform if the computing 

power of an ARM9 becomes inadequate. 

This is available as a finished, check-card sized 

core module with all important components, 

and can be used on a customer-specific baseboard 

that is yet to be developed. Individually 

configured, it contains all the coarse components, 

such as sockets, plugs, transmitters and 


Figure 1. Example of 

a customer-specific 

solution based on an 

ARM9 module with 

a 5.7“ QVGA and an 

8.4“ VGA touch display, 

ready for installing in 

a metal housing 

power supply. The ARM11 module is also 

available with embedded Linux or Windows 

CE. The ARM11 not only offers many times 

the computing power of an ARM9, but can 

also resolve graphics up to XGA format and 

show videos in real time. For the top class, 

EDM still has Power PC platforms in its range, 

which are used to perform data-intensive and 

highly graphic tasks. In this case, not only embedded 

Linux but also VxWorks can be offered 

as the real-time operating system. 

The ARM9 processor platform is available in 

a low-cost version with a graphics-capable display 

(64x160) and Jog-Dial as the input medium. 

The pure processor module is available 

for use in individually prepared solutions. 

Housing, display, input medium and an interface 

module can be added according to customer 

requirements. There are even readymade 

solutions for high-end requirements. As 

SITOUCH, the ARM9 platform can be supplied 

with either a 5.7“ QVGA or a VGA display and 

touch in a flush wall housing, or with an 8.4“ 

VGA display and touch in a stand-alone metal 

housing. These platforms all run with 24 VDC, 

without a fan, and in an ambient temperature 

up to 60°C. The EDM concept is highlighted 

by the fact that the platform, which is also 

known as main modules, can be coupled via a 

proven, standard RS485 interface to one or

Figure 3. Example platform of a check-card 

sized ARM11 module 

Figure 4. Customer-specific sensor-actuator 

module with expansion modules plugged in 

more customer-specific sensor-actuator modules. 

By means of a proprietary protocol, this 

interface ensures that the intelligence of the 

controller is informed at all times of the 

current status of the entire machine and can 

affect it in real time. 

Two-module solutions have proven themselves 

in several projects, each consists of an ARM9 

HMI module adapted for the project, and a 

sensor-actuator module developed individually 

for the project. In line with the EDM philosophy, 

the sensor-actuator module is composed 

of previously tried and tested basic I/O configurations, 

and controlled and driven by an 

ARM7 or Cortex M3. The ARM7 or Cortex 

M3 communicates continuously with the 

ARM9 head via the serial interface and the 

special machine protocol. The concept is designed 

so that either multiple sensor-actuator 

modules can be cascaded in order to realize, 

for example, different expansion stages for different 

types of machine. Or individual expansion 

modules can be plugged into interface 

sockets on the sensor-actuator module. These 

are, in turn, automatically recognized and integrated 

by the ARM7, and finally by the operating 

system. This example can be realized 

equally well with ARM11 or Power PC instead 

of the ARM9. The requirements of the customer 

system are decisive. n 


Figure 2. Example of a customer-specific solution in the low-level segment consisting of an ARM9 

head assembly with embedded Linux (a) and an ARM7-based sensor-actuator module (b) 

21 September 2010


Embedded designers learn their 

lessons from the battlefield 

By Jens Wiegand, Wind River 

Commercial software vendors 

are combining their experience 

in various industrial segments 

with the lessons of the military 

and aerospace industries, 

leading to higher-quality 

system designs without drastic 

increase in project time or 

cost, and allowing for the use 

of legacy and open source 

code with virtualisation 

software to separate 

safety-critical functions. 

n Designers of industrial equipment are learning 

from other markets to take advantage of 

the latest technologies without having to go 

through long certification processes and shortening 

the time-to-revenue. Safety and reliability 

are increasing concerns when designing industrial 

equipment, from the factory floor to 

trains to medical equipment. As the pressure 

on development time and cost increases, designers 

are looking for ways to make the 

processes more efficient; but at the same time 

they want to add new features and conform to 

new homologation or safety standards. Emerging 

standards such as CENELEC EN50126, 

EN50128, EN50129, and the European Train 

Control System (ETCS) are making design 

times longer and less predictable because more 

certification needs to be done, with different 

requirements for different countries. With new 

communication systems in transport and automation 

systems, certification is more complex 

and time consuming. 

This is a complex combination of requirements 

that can no longer be met with the traditional 

approach of custom hardware and software, 

so commercial off-the-shelf (COTS) solutions 

are increasingly being adopted. At the same 

time there are millions of lines of code that 

have been well proven in the field over many 

years, so designers want to be able to include 

this legacy code in new developments. Using 

pre-tested and proven functions is a key way 

to reduce the design time and the time-torevenue 

of a complete system. One of the key 

trends emerging across many industrial segments 

is the need for time and space separation 

of functions. Traditionally this has been 

provided through separate hardware boards 

running separate software stacks; but this is 

increasingly costly to develop and maintain. 

Now it is possible for uni-core and multi-core 

processors coupled with new virtualisation software 

to separate the safety-critical functions, 

allowing legacy code to run on one processor 

with non-critical code running on another, all 

under the control and protection of a robust, 

lightweight hypervisor. This virtualisation 

model has been increasingly popular in the 

aerospace and military industries. These designs 

have moved to COTS boards and separation 

kernels over the last few years, and these systems 

are now emerging into front-line service. 

Of course, these systems have to conform to 

highly specialised reliability standards such as 

DO-178 and ARINC 653, which means that 

developers must demonstrate high levels of 

confidence in system performance. This involves 

a multitude of test cases and test artefacts 

and applied tools, often provided by commercial 

software suppliers such as Wind River and 

its ecosystem of partners, to demonstrate the 

high levels of safety inherent in the system design 

and implementation. The developments 


are also optimised for the long-term needs of 

markets where the same designs can be in use 

for 10 or 20-plus years. The same techniques 

are now being adopted in industrial standards. 

Embedded software suppliers can use their experience 

in high-reliability applications to bring 

high levels of safety to the industrial market 

without the penalty of long design times and 

costly implementations. Using a commercial 

real-time operating system in combination 

with a hypervisor on mainstream uni-core or 

multi-core processor boards provides design 

flexibility. Meanwhile, the pre-assessed artefacts 

and test cases that run on these platforms for 

the industrial segments help to speed up both 

development time and the certification process. 

This avoids costly reruns and bug fixes, moving 

an inflexible hardware-centric product lifecycle 

process to a highly flexible software 

oriented life-cycle process. 

This allows modularity, reusability, and maintainability, 

which in turn allows device manufacturers 

to focus on innovation and faster 

time-torevenue. Networking is an increasingly 

important part of any industrial design. Equipment 

on the factory floor is being linked together. 

Similarly medical equipment has to be 

able to communicate with other systems in 

the hospital or in the home. In both cases it is 

vital that communications are reliable and secure. 

However, networking stacks are notorious 

for being a point of external access to a system.

Figure 1. Illustration of the use of safety-critical 

solutions. A certifiable hypervisor is 

used, on which a certifiable application 

under either VxWorks, or even without an 

operating system, is running in one partition. 

The application in the other partition is not 

subject to any restrictions critical to safety 

and is a Linux with graphical visualization 

in this example 

Figure 2. Flexible expansion of existing 

applications with simultaneous hardware 

migration. The existing implementation is 

reused and new functionalities will be added 

in a second hypervisor partition 

Figure 3. The graphic illustrates the distribution 

of the physical hardware to virtual 

boards - in this case three. The operating 

systems Linux, VxWorks and, at the rightmost 

end, an application without an 

operating system are running on them. 

MIPC denotes a protocol for fast intrapartition 

communication 

The communication also has to be securely 

transmitted to prevent snooping and intrusion, 

particularly when using Ethernet or connecting 

to the wider internet. The control industry is 

facing an increasing amount of cyber attacks, 

shifting security scrutiny from the finance and 

defence industries to the control industry. 

Even with an established fieldbus protocol, 

separating the network stack can enhance reliability. 

It also allows communications to be 

quickly and reliably upgraded to higher speeds 

or new versions of the network standard, without 

having to rewrite the rest of the system 

code. As a result, software vendors are providing 

certified network stacks that run on high-performance 

multi-core processors from companies 

such as Intel, Freescale Semiconductor, 

Cavium, and NetLogic. This alone is not 

enough to achieve certification, as the whole 

system has to be certified. Nonetheless, a certifiable 

stack running on an isolated core with 

clearly defined communication paths can dramatically 

reduce the certification time. Even if 

the communications code crashes or is compromised, 

the rest of the system is protected 

and the affected processor core can be restarted 

without having to implement a global reboot 

or, worse, field servicing by a technician. 

The increased computing power of multicore 

processors and their improvement in 

performance, power, and price over previous 

uni-core processors has created the possibility 

of consolidating various disparate systems 

onto a single device. Leveraging innovations 

in virtualization technology such as high-performance, 

deterministic hypervisors, multicore 

processors can host real-time safety-critical 

systems and general-purpose operating 

systems. Combining different functions and 

different levels of criticality on one device 

with time and space separation is a key paradigm 

shift. Vehicle control units (VCU) or 

human-machine interface (HMI) platforms 

combined with safety functions are good examples 

of this shift. As a result, the control 

and user interface functionality can be combined 

or consolidated on a single platform, 

and the HMI can be implemented on a core 

www.embedded-control-europe.com 

The Online Information Source for Europe`s Europe s Embedded Engineers 

• Product Product News 

News • 

TTechnical 

echnical Articles 

• 

Com Company mpany p y Profiles 

• 

Events 


that is entirely separate from the control and 

even safety-critical elements of the design. 

This allows for new features, functions, and 

interfaces such as 3D graphics without a major 

recertification of the entire system. It also 

makes the equipment easier and safer to use. 

The time and space separation further allows 

users to employ general-purpose operating 

systems (GPOS) such as Linux on a separate 

core from the safety-critical code. This allows 

designers to access a broad ecosystem of software 

that they can modify for their own applications 

without compromising the core functionality. 

It speeds up development and helps 

to reduce costs. This combination of multicore 

processors and commercial software offers 

considerable flexibility in safety-critical designs 

for the transportation or energy markets. 

For example, in a vehicle control unit, one core 

can be running a hypervisor with a real-time 

operating system while another can be running 

a bare metal implementation of a small safety 

application or a more complex safety application 

using a certified real-time operating system. It 

prevents the same root cause of failure from 

compromising the whole system. 

There are a number of reasons why this approach 

is being adopted by companies such as 

Bombardier, Alstom, and Siemens for transportation 

systems. Firstly, transportation systems 

are often large contracts that span many 

years. There are extensive penalties if these 

projects overrun, as well as increasing pressure 

on costs and schedule for the existing contracts. 

Developers can no longer take many years to 

bring a system to market nor can they overspend 

the budget with impunity. Secondly, 

new train and tram developments are also 

using more and more electronics, including 

automatic train protection systems, automatic 

train operations, and driver and passenger information 

systems. On top of these are the 

new emerging video and internet data services 

being offered to passengers, which drive convergence 

of control and passenger comfort information. 

The train makers are exposed to 

and simultaneously challenged by this wealth 

of new requirements. 

Both these factors are now driving the use of 

COTS hardware and software rather than 

companies developing these technologies inhouse. 

They need to meet the specific needs 

of the transportation market in reliability, 

long-term support over the 20 or 30 years of 

the equipment life, and compliance with safety 

regulations. As virtualisation and multi-core 

devices have moved from the PC arena to the 

embedded world, the opportunities for increasing 

the safety and efficiency of industrial 

designs have increased. n 

23 September 2010

SMALL FORM FACTOR BOARDS 

PICMG adopts new specification 

for COM Express modules 

By Gerhard Szczuka, Kontron 

COM Express is extremely 

important in the embedded 

market as the only vendorindependent 

standard for 

Computer-on-Modules, other 

than the one adopted in 1998 

for DIMM-PC and the one 

published in 2000 for ETX. 

With Rev.2.0 it has been 

extended with a new form 

factor and new pin-outs to 

meet future needs. 

n Five years after its debut, the PCI Industrial 

Computer Manufacturers Group (PICMG) 

has now released revision 2.0 of the COM Express 

COM.0 Computer-on-Module standard. 

COM.0 Rev.2.0 addresses the new functionalities 

that Intel, AMD, and other manufacturers 

are integrating in their upcoming processor 

families. Rev 2.0 adds two new pin-outs that 

make space for possible future technologies 

by dropping a number of interfaces that are 

less useful. The growing significance of graphics 

and displays is evident in the changes made to 

the COM Express specification. These optimizations 

now make smaller COM Express 

form factors possible. Even with the changes 

to the specification, the pin-out types produced 

to date will continue to be available with new 

generations of chips, thereby ensuring the scalability 

of existing applications. 

In the COM Express COM.0 specification, the 

PICMG defines the standard for a Computeron-Module 

(COM) as a bootable host computer 

in the form of a single large-scale integrated 

component. The vendor-independent 

specification of interfaces and form factors 

for Computer-on-Modules gives designers and 

solution providers a firm basis on which to develop 

products for the market that are futureoriented 

and promise long-term availability. 

PICMG continues along the right path with 

the Rev 2.0 of the COM Express specification. 

OEMs that set up their medical diagnostic apparatus, 

industrial robots and vending machines, 

test and measurement applications, 

POS and kiosk systems, surveillance cameras, 

or unmanned vehicles on COM Express modules 

have chosen a sustainable and innovative 

solution for the future. When these designers 

elect to migrate to a more compact form 

factor, COM Express COM.0 Rev.2.0 will afford 

them a seamless transition, thanks to the newly 

specified compact form factor (95mm x 

95mm). Kontron has offered the compact 

form factor under the registered trademark 

microETXexpress since 2006. 

A change that is a little less apparent but that 

has a great impact is the addition of two new 

pin-outs to the five already defined pin-out 

types within the COM Express specification. 

Seven pin-out types are now defined in Rev.2.0. 

These can be split into two groups differing 

initially in the number of connectors they utilize. 

Pin-out types 1 and the new pin-out type 

10 utilize the single A-B connector that has 

220 pins, which can also be found on all other 

pin-out types. But types 2, 3, 4, 5, as well as 

the new type 6 also use the second 220-pin 

connector - the C-D connector - so they 

possess a total of 440 pins. Let us now look at 

the features of the individual pin-out types as 

specified in COM.0 Rev.2.0. Pin-out type 1 

has one 220-pin connector - the A-B connector 


Figure 1. The Kontron COM 

Express COM.0 Rev.2.0 Type 6 

ETXexpress-AI with Intel Core 

i5/i7 Processor 

- and supports up to eight USB 2.0 ports, up 

to four SATA or SAS ports, and up to six PCI 

Express Gen1/Gen2 lanes. It supports dual 

24-bit LVDS, an HDA digital audio interface, 

Gigabyte Ethernet, and eight GPIO pins. SPI is 

added to all single pin-out types in Rev.2.0 on 

previously reserved pins. (We will go into this 

issue in more detail below.) The primary input 

voltage is +12V and standby is +5V. Some solutions 

such as Kontron COM Express module 

allow a variable input voltage. Pin-out type 2 

has all the stated functionality of type 1, but 

adds a second 220-pin connector to it as well 

(C-D). In this case, type 2 features a 32-bit PCI 

interface plus IDE ports to support legacy PATA 

devices such as PATA HDD and CompactFlash 

memory cards. There are a total of 22 PCI 

Express lanes (six on the A-B connector and up 

to 16 on the C-D connector), 16 lanes on the 

second connector being intended for PCI Express 

Graphics (PEG). The maximum power 

consumption, previously defined as 188W, is 

now matched to 137W in Rev.2.0, thanks to 

ever-more energy efficient processors. 

Comparing pin-out type 3 to type 2, only the 

IDE pins are used in favor of extra Gigabit 

Ethernet capability. Consequently, it has no 

legacy interfaces, but now supports up to three 

Gigabit Ethernet channels. In pin-out type 4, 

again compared to type 2, pins reserved for 

PCI are reallocated, creating space for ten


Figure 2. The new COM Express compact form factor (95mm x 95mm) opens the way for 

compact designs within the COM Express specification. 

additional PCI Express lanes. These can be 

used as PCIe lanes 0-15 or as second PEG port 

lanes 16-31. Pin-out type 5 fuses the changes 

in types 3 and 4 as compared to type 2. Pinout 

type 6 opens up a new world of graphics. 

PICMG has added a sixth type of pin-out to 

the COM Express standard especially to utilize 

the expanded graphics possibilities of new 

processor families. This pin-out type is essentially 

based on type 2, the most widely adopted 

COM Express COM.0 pin-out type to date. 

Legacy PCI pins are now used to support the 

digital display interface and for additional PCI 

Express lanes. Furthermore, in pin-out type 6, 

the pins previously assigned to the IDE interface 

in pin-out type 2 are now reserved for future 

technologies still in development. One of these 

technologies could well be SuperSpeed USB, 

because the 16 free pins would offer sufficient 

lines to implement four of the eight USB 2.0 

ports as USB 3.0 ports that each requires an 

extra pair as compared to USB 2.0. Although 

pin-outs type 2 and 6 are very similar, pin-out 

type 6 adds extensive support for additional 

display interfaces. The graphics options have 

always been one of the special strengths of 

COM Express through support of PEG, chosen 

by Intel to provide a high-speed bus for external 

graphics cards. But these days it is not only a 

matter of satisfying the growing need for performance, 

but of supporting different output 

devices as well. And these are precisely the 

kind of requirements for which pin-out type 

6 has been scaled. Like virtually all the other 

pin-out types (except type 10), pin-out type 6 

continues to support the familiar analog VGA, 

the standard interface for RGB/CRT devices 

used in many industrial applications. Because 

of the analog transmission of the picture signal, 

VGA is not entirely suitable for modern TFT 

displays with resolutions of more than 

1280×1024. LVDS (low-voltage differential signaling) 

devices can also be driven directly by 

all pin-out types. This is important for applications 

with LCD displays, for example, which 

mainly use this transmission standard. Here it 

should be noted that the dual 24-bit LVDS 

channels are designed for one display; the second 

channel serves solely to process the increasing 

data rates caused by higher resolution 

and frequency. The connected LVDS display 

defines how many channels are needed for 

each resolution. 

Pin-out type 6 goes far beyond these graphics 

options. It offers three new ports that are dedicated 

to new digital display interfaces (DDI). 

The developer can configure these ports individually 

for HDMI (high-definition multimedia 

interface), or the electrically compatible DVI 

(digital visual Interface) or DisplayPort (DP). 

DDI port 1 supports additionally SDVO (serial 

digital video output). SDVO is not multiplexed 

on the PEG port in type 6, which has been 

possible under the pin-out type 2 definition. 

Thus, in parallel with embedded graphics, an 

external PEG graphics card can be used, for 

example, for multi-monitor applications with 

more than 4 screens or for data processing 

using general purpose GPUs (GPGPU). With 

the SDVO interface supported by Intel chipsets, 

COM Express is flexible in supporting a wide 

variety of graphics signals. So the developer 

25 September 2010


can now implement DVI (digital visual interface), 

for example, and achieve relatively lowcost 

integration of digital monitors and dual 

display solutions. COM Express previously 

did not officially support SDVO, but it has 

also become part of the COM Express standard 

in the new COM.0 Rev.2.0. Graphic layout: 1: 

VGA, 2: LVDS, 3: DDI (SDVO;DP; HDMI 

(TMDS)), 4: DDI (DP; HDMI (TMDS)), 5: 

DDI (DP; HDMI (TMDS)). 

The developer can also operate modern DisplayPort 

and HDMI/DVI graphical interfaces 

through the DDI. DisplayPort is a universal 

and - unlike HDMI - royalty-free interconnect 

standardized by VESA, which should ensure 

its widespread popularity. DisplayPort not 

only has a much higher data transfer rate of 

17.28 Gbps (compared to 2.835 Gbps with 

LVDS and 4.95 Gbps with DVI), but also a 

micro-packet protocol, allowing simple expansion 

of the standard. Furthermore, DisplayPort 

supports an auxiliary channel that allows a 

bidirectional connection to control devices by 

VESA standards such as E-DDC, E-EDID, 

DDC/CI, and MCCS. This enables genuine 

plug and play operation. The auxiliary channel 

can be used for peripherals such as touchpanel 

displays, USB connects, cameras, microphones, 

etc. DisplayPort could eventually replace 

HDMI, popular on the consumer market, 

which as mentioned already is also supported 

by the COM Express standard on the DDI. 

HDMI is an ideal solution for AV and multimedia 

applications, such as home theater PCs 

or set-top-boxes, due to its high data rate, its 

connector concept (audio and video on one 

cable), and its backward compatibility. However, 

this interface was not developed for the embedded 

market. Its implementation does not 

make for a particularly stable solution, and 

long-term availability could also be a problem 

as drivers or mechanical requirements frequently 

change. 

With this extensive support for the new graphics 

and display functionalities of upcoming 

chipsets, pin-out type 6 is a promising followon 

to pin-out types 2 and 3 and comes at the 

right time. Kontron already anticipated this 

development in a number of areas, so developers 

who want to make full use of the new 

Table 1. With Type 10, the SATA 2 and 3 pins are no longer occupied. They are now reserved for 

alternative purposes, such as USB 3.0. 

Table 2. In Type 10, the pins for PCIe Lanes 4 and 5 remain free and can be used for future 

technologies. 

Table 3. COM Express COM.0 R2 supports two independent displays via LVDS and DDI. 


graphics possibilities of COM Express are well 

served by the embedded specialist. Given its 

experience, Kontron can provide developers 

with optimal support when migrating from 

pin-out types 2 or 3 to pin-out type 6 and 

help to ensure a seamless transition. So, with 

the ETXexpress-AI, also one of the first type 6 

modules in the basic form factor is presented. 

The major innovation that COM.0 Rev.2.0 

represents is the definition of the new pin-out 

type 10, a kind of twin brother to pin-out 

type 1. Type 10 addresses the requirements of 

newer and highly compact processors more 

explicitly. A close look at the pin assignments 

reveals the differences to watch out for when 

migrating from type 1 to this new type, although 

both pin-out types are compatible 

with each other. In pin-out type 1, SATA ports 

2 and 3 are assigned pins in rows A and B, but 

these are no longer reserved in pin-out type 

10. The pins could still be used as SATA ports, 

but are now reserved for alternative purposes 

such as USB 3.0. So in designs for pin-out 

type 1 as for type 10, Kontron advises against 

wiring SATA 2 and 3 over the module connector. 

The modules then remain compatible, and 

they are ready for USB 3.0 at the same time. 

Table 1 indicates the differences and shows another 

difference in rows A and B, this time regarding 

the pinning of the PCIe lanes, where 

pin-out type 1 offers six lanes in total. In pinout 

type 10, the pins for PCIe lanes 4 and 5 

are no longer reserved and can also be used 

for upcoming technologies. The background 

in both these cases is as follows: processors of 

a small form factor, at which type 10 aims, 

support up to two SATA interfaces and four 

PCIe lanes. The vacated pins on the module 

connectors of the ultra standard can therefore 

be used efficiently for new purposes. As of 

COM.0 Rev.2.0, serial ports are again supported. 

The pins for this were previously used for 

VCC 12V. 

What is new is that with COM.0 Rev.2.0 Type 

10 and Type 6 support serial ports. The pins 

were formerly used for VCC 12V. However, 

dedicated manufacturers ensure compatibility 

with existing carrier boards by a protective circuit 

on the module. Developers do not have 

to completely modify their existing carrier 

board layout, but can cost- and time-efficiently 

use the new possibilities. A further difference 

is that type 10 uses the second LVDS channel, 

TV out and VGA to support the SDVO port 

(or alternatively DisplayPort or HDMI/DVI) 

via DDI. That is no real loss seeing as VGA 

will only play a minor role in future. But now 

type 10 ultra-compact modules (such as the 

nanoETXexpress-TT) provide native support 

not only for latest display interfaces but also 

for dual independent displays, since they will 

continue to support an LVDS channel. Table 3

Table 4. The changes from type 1 to type 10 at a glance. 

shows the precise differences in pinning between 

types 1 and 10. For customers that already 

use the ultra compact nanoETXexpress 

modules these differences are not likely to be 

of much consequence: With foresight, Kontron 

has already reserved the appropriate pins for 

SDVO support, for instance, on the former 

VGA and second channel LVDS pins in its nanoETXexpress-SP 

modules. 

Further changes affecting all the types of modules 

available are as follows. The COM Express 

connector in the present form is now also approved 

for PCI Express Gen2 signals. Technically 

speaking, that means no alteration to the 

connector or its pinning, but the developer 

must still adhere to new rules for PCIe Gen2 

when routing the module and carrier board. 

Additionally, the AC97 pins are now used to 

support AC97 and HD audio. Kontron users 

will already be familiar with this as the majority 

of Kontron COMs support these audio features. 

The following changes have additionally been 

made in the new version. COM modules 

type10 and type 6 now also support SDIO, 

multiplexed on the existing GPIO signals. Optionally, 

two 3.3V TTL serial ports are added - 

as required by many legacy applications - and 

here the standard again shows its flexibility in 

responding to the needs of the market. Both 

ports can be used for different purposes such 

as RS232, RS485, the CAN bus, or other twowire 

interfaces. 

One change shown in the new specifications 

affects all pin-outs: in addition to the previous 

firmware hub, there is a new BIOS or firmware 

interface for an internal and external boot im- 

plemented in the new generation of processors. 

This is a serial peripheral interface (SPI), the 

future interface for firmware flash on the module 

and carrier board. Ready reserved pins are 

used for this purpose. Generally, PICMG allows 

a choice between two SPI chips, the new 

COM.0 Rev.2.0 specifying external firmware 

support for all module types. The LPC interface 

was used for this purpose in the earlier version. 

The new modules must support SPI, but may 

still additionally flash firmware externally 

through LPC, i.e., if the chipset continues to 

support it. The reason for this change in 

firmware flash is that the new small form 

factor processors only support SPI boot devices. 

Inclusion of the smaller compact form factor 

in the standard is a major innovation. It means 

that the most widely adopted pin-out type 2 

can now also be used in applications with 

space constraints. COM.0 Rev.2.0 defines its 

dimensions as 95mm x 95mm. Aside from the 

reduced footprint, the physical requirements, 

connector placement, and pin-out are exactly 

the same as those of the successful basic form 

factor. This is another case where Kontron entered 

the scene early, because the company 

had already been producing modules with 

these specifications for more than two years 

under the brand name of microETXexpress, 

and was the first one to market. Recently, it introduced 

a new product in this module family: 

the microETXexpress-XL with an Intel Atom 

Z520PT processor and Intel US15WPT system 

controller hub. This is a COM Express COM.0 

pin-out type-2 Computer-on-Module with a 

compact form factor, specially developed for 

use in the E2 industrial temperature range 


from -40 to +85°C. The microETXexpress-XL 

with a 1.33 GHz Intel Atom Z520PT processor 

supports up to 2 gigabytes of soldered DDR2 

RAM, and also has space for an onboard solidstate 

drive. Additionally, it makes full use of 

the bandwidth of the COM Express pin-out 

type 2 connector with 1x Gigabit Ethernet, 1x 

serial ATA, 1x PATA, 8x USB 2.0 and 2x PCI 

Express, plus PCI for custom additions. With 

the SDVO port, it is very simple to implement 

a DVI output, and together with the 24-bit 

LVDS single channel the microETXexpress- 

XL presents possibilities for connecting a whole 

variety of displays and monitors. This extensive 

feature set makes it one of the most flexible interface 

choices in Computer-on-Modules in 

the compact COM Express form factor for extreme 

environments (-40 to +85°C). 

The latest example of the microETXexpress- 

XL shows that Kontron responds to customer 

requirements at an early stage. The case is a 

similar one with the even smaller modules of 

the nanoETXexpress family (84mm x 55mm), 

for which a large demand exists on the market. 

nanoETXexpress modules also match the 

PICMG COM Express standard with respect 

to pin assignments and pin-outs type 1 and 

type 10 and are well suited for a new generation 

of mobile embedded applications with low 

power consumption and the latest interfaces 

no bigger than a credit card. n 

Product News 

n Fastwel: low-voltage PC/104-Plus SBCs 

and peripheral modules 

Two new Fastwel PC/104-Plus single board 

computers â€“ CPC306 and CPC307 are 

available for ordering now. Fastwel CPC306 

is designed for applications requiring industrial 

controller with low power consumption 

and compact size but capable to run in 

industrial environments. It is based on 

DM&P Vortex86DX (600 MHz) processor, 

with 256 MB DDR2 SDRAM, integrated 

Ethernet channels, four COM ports and 

four USB 2.0 ports. 

News ID 11040 

n Moxa: ARM9-based computers with 

2 serial ports and dual LANs 

The IA3341 is based on the MOXA ART 

ARM9 industrial processor, and features two 

RS-232/422/485 serial ports, dual LANs, four 

digital input channels, and four digital output 

channels. In addition, the IA3341 computer 

has two analog input channels and two thermocouple 

channels, making it an ideal solution 

for a variety of industrial applications. 

News ID 10994 

27 September 2010


Replacing discontinued modules forces 

ETX module transplantation 

By Konrad Löckler, MSC 

The following article provides 

information on points to 

consider, such as long-term 

availability, in order to ensure 

that an ETX module 

transplantation is as simple as 

possible. 

n Over the course of many years, system manufacturers 

that relied on the use of ETX modules 

had little reason to change anything on 

their established devices. Reliable ETX products 

were available unchanged over a long period 

of time. However, since Intel’s notification of 

discontinuance of the widely used Pentium M 

and Celeron M generations, development departments 

have been feverishly discussing how 

to handle this situation: a new design of the 

entire platform, or replacement of the ETX 

module with a current model? 

Manufacturers from various sectors must equip 

their successful product lines with a current 

processor core as smoothly as possible. Because 

the change to another form factor of embedded 

COM modules involves a great deal of time 

due to the effort required for integration, and 

therefore such a step makes more sense when 

developing a completely new product generation, 

ETX suppliers are receiving an increasing 

number of enquiries regarding successor products 

that have long-term availability. With industrial 

controls, operating panels, robot controllers, 

studio mixing consoles, textile machines, 

measuring instruments or medical 

equipment, the question always arises whether 

replacement of the processor module can actually 

be made as smoothly as possible, as the 

module suppliers always claim. The most current 

and thereby longest available ETX modules 

are based on Intel Atom technology. Two 

performance classes are thus fundamentally 

possible: single core solutions with 1.6 GHz, 

which include the Intel Atom N270 based 

MSC ETE-A945GSE, and dual-core modules 

with 1.66 GHz such as the MSC ETE-PV510, 

which is based on the Intel Atom D510. Measurements 

performed in-house at MSC show 

that the computing performance of an N270based 

product approximately corresponds with 

the computing power of a Celeron M with 1.3 

GHz. Therefore, if products with Celeron M 

from 600 MHz to 1.3 GHz have so far been 

used then an N270 processor should be 

considered. 

For higher computing power requirements, a 

product with the D510 processor with two 

cores and multi-threading offers performance 

that is practically double, provided that the 

software can still be used with these newer 

technologies. As experience shows, a dual-core 

processor under Windows brings performance 

increases of up to 30%, even with applications 

which run in only one thread. The reason for 

this is that at least the operating system and 

driver can run on the second core, and the 

application program has a core to itself. Adaptation 

of the operating system installation is 

unavoidable when replacing the processor technology 

in a system. Network drivers, graphics 

drivers and some other drivers must be replaced 


Figure 1. In the medium performance 

range, the power saving 

ETE-A945GSE with Intel 

Atom N270 offers a long-term 

availability alternative to 

existing Celeron M modules. 

in order to be able to use the new controllers. 

The widespread Atom architecture offers the 

significant advantage that the necessary drivers 

are often already integrated in the most popular 

operating systems, such as Windows and many 

Linux distributions. What is lacking is normally 

made available by the hardware manufacturers. 

MSC also offers broad support with the integration 

of Windows CE and other operating 

systems. Many important module functions 

and internal information, - such as querying 

serial number, boot counter or BIOS version, 

access on watchdog, temperature values, EEP- 

ROM, I²C bus or brightness control of the display 

- are provided via a uniform program interface 

with the products presented here. The 

associated drivers for Windows, Windows CE 

or Linux are available from the supplier. 

The necessary or available display interfaces 

also need to be carefully considered. In the 

simplest case, a monitor is driven with a VGA 

connection. This always works and with Atom 

modules even ranges from 640 x 480 pixels 

right up to a resolution beyond full HD of 

2048 x 1536 pixels. However, embedded systems 

normally contain TFT displays with various 

interface technologies. With LVDS, care must 

be taken that 18-bit displays can be easily supported, 

but not always 24-bit panels. The most 

flexible here is the MSC ETE-A945GSE module 

that can even drive two LVDS displays; one

Figure 2. Computing power of current Intel Atom processors compared to Pentium M and 

Celeron M generations, measured on MSC ETX modules with SiSoftware Sandra 2007 

Benchmark 

with 24-bit and a further one with 18-bit color 

depth. The maximum resolution is 1600 x 1200 

pixels. The second LVDS interface is located on 

a connector directly on the ETX module and 

can be used with a special cable. With the MSC 

ETE-PV510, due to the different chipset, an 18bit 

interface with up to 1366 x 768 pixels is supported. 

The most popular display formats are 

offered via selection tables in the BIOS setup 

program and can be used immediately. Less 

common resolutions can normally also be controlled 

via a suitable EDID parameter set, which 

is stored in an EEPROM. The manufacturer offers 

tools and support for this. 

ISA bus based controllers or plug-in cards are 

often still used with systems that have already 

been on the market for several years. With the 

PCI-to-ISA bridge used, the new ETX modules 

offer a complete ISA bus, which is fundamentally 

compatible to existing ISA environments. 

However, one limitation exists when DMA access 

via ISA are necessary. Intel no longer supports 

the relevant DMA protocol with platforms 

after the 855 chipset generation. An 

early clarification with the module supplier 

regarding possible workarounds is recommended 

in cases where DMA operation via ISA still 

is required. Since the introduction of the ETX 

specification 3.0 several years ago, ETX modules 

usually have two SATA connectors in addition 

to the two IDE (PATA) channels. Modern 

drives can be directly connected to these SATA 

connectors. In the event that IDE drives - 

which are becoming increasingly difficult to 

acquire - are still used in the system, the 

chance is offered here also to replace them 

with SATA devices. Both the ETX modules 

from MSC that are recommended offer - in 

addition to four USB 2.0 interfaces routed to 

the baseboard - two further interfaces, which 

can be tapped on the module via mini-USB 

connectors. Transmission problems with USB 

high-speed connections, which in the past occasionally 

occurred with ETX, can thereby be 

ruled out. The optional availability of a trusted 

platform module (TPM) was also previously 

not common. For safety-critical applications, 

the possibility to use such a TPM is now 

offered in order to prevent unauthorized manipulations 

of the system or to encrypt data. 

For a long time now, CompactFlash cards 


have been popular in embedded systems as robust 

hard disk replacements. Meanwhile, newer 

flash formats exist and especially with ETE- 

PV510 a version with soldered silicon disk, 

from which the operating system can be booted, 

is offered. 4 Gbytes is the most popular size; 

however, capacities up to 16 Gbytes are also 

possible. If a module equipped with silicon 

disk is used, it should be noted that the first 

IDE channel is thus occupied and only the second 

IDE channel (master/slave capable) remains 

available for further drives. A final mention 

must be made that the D510 processor 

on the ETE-PV510 is also 64-bit capable (Intel 

64 architecture) and offers four threads on the 

two processor cores. The N270 with two threads 

also supports Intel hyper-threading. 

In addition to software compatibility and electrical 

characteristics, heat dissipation must be 

considered when changing to a more powerful 

module. Because suitable Atom processors are 

consistently implemented in energy-efficient 

technologies, a power dissipation of approximately 

8 watts can be expected with MSC 

ETE-A954GSE and approximately 15 watts 

with ETE-PV510. If the older modules that 

need to be replaced produced similar or higher 

heat losses, a good contact to the already existing 

cooling measures must be provided during 

replacement for the so-called hotspots. These 

hotspots are generally the processor and the 

chipset. Where a standard ETX heat spreader 

has so far been used, such a heat spreader is 

also available for the new modules. Thus, a defined 

interface for transferring heat exists 

which can be mounted in the same way as before. 

Alternatively, passive or active cooling solutions 

are also offered which do not require 

any further measures, because the heat is safely 

extracted by means of cooling fins and 

temperature-controlled fans. n



exist alongside the x86 standards? 

By Wolfgang Heinz-Fischer, TQ-Group 

This article reviews the 

proliferation of standards 

following x86, so that it is now 

difficult to define a standard in 

the clear x86 market and even 

more so in the non-x86 

market, and argues that 

proprietary systems have a role 

in the non-x86 market, 

providing the full functionality 

of the chosen processor 

on one hand and the 

interchangeability of x86 

systems on the other. 

n Standardization of x86 systems for industrial 

applications dates back to the nineties, when 

the first PC/104 systems and ISA slot cards 

were introduced. Standardization here refers 

to the form factor and the joint bus system, in 

this case the ISA bus. The advantage for the 

user was readily apparent in the large number 

of exchangeable or complementary systems 

by various vendors. 

Over the years, additional bus systems, such as 

the PCI bus and now PCIe bus, were added, 

leading to new standards. In early 2000, ETX 

was introduced to the market as the first standard 

module with a defined range of functions. 

Not only the ISA and the PCI bus are specified 

here, as extensions to PC/104, but additional 

PC functions are realized via the plug-in system. 

The most recent version of a corresponding 

system with PCI and PCIe bus, COM Express, 

was introduced in 2004. The currently valid 

standards are ETX, XTX, Q7, COM Express, 

PC/104, 3.5“ SBC, EPIC, PICMG1.0, PICMG1.3 

and all types of industrial mainboards from 

ATX to MiniITX. These standards are administered 

either permanently under organizations 

such as PICMG of PC/104 Org. or permanently 

defined within workgroups or stakeholders. 

Standardization is possible in this case because 

the x86-architecture specifies uniform interfaces, 

although parallel buses have been replaced 

by faster serial buses in recent years. 

Thus, for PC or x86 standards, PCIe, USB, 

Fast or Gigabit Ethernet, SATA, DVI, LVDS, 

audio, mini PCIe and SD card are the rules 

today. The PCI bus is losing market share. 

If, however, the application is closer to a direct 

machine control, different interfaces and functions 

play a role. Serial interfaces, GPIOs and 

field buses are important here. In the PC 

world, these interfaces are implemented 

through in part elaborate additional cards. 

This is exactly where ARM-based processors 

apply - the combination of PC-based basic 

functions coupled with special machine-oriented 

interfaces. The significantly lower power 

loss is invoked by processor manufacturers as 

an additional criterion. The diversity of processor 

variants outside the x86-world for which 

module solutions and so-called standards are 

offered, is almost limitless; starting with 16bit 

microcontrollers, such as Infineon C166 

family, through various ARM derivatives in 

various performance categories, and ending 

in the upper range with complex PowerArchitecture 

processors such as the PowerQUICC 

III or QorIQ by Freescale. In this sector, the 

number of CPU manufacturers has a very 

broad base, reaching from A as in Atmel to Z 

as in Zilog. The performance category from 8bit 

to 64-bit, and likewise the markets and applications 

addressed, have a correspondingly 

broad base. Thus, is a standard, as offered by 


Figure 1. Back side of the 

embedded module TQM5200 

several manufacturers, even possible here? And 

what actually is a standard? A true standard is 

a firm definition of the functions and the mechanical 

values of a processor platform that 

are administered and registered in a non-profit 

organization such as PICMG, among others. 

True standards are designed to ensure that devices 

manufactured by different vendors are 

interchangeable. In addition, there are quasi 

standards that are supported jointly by a larger 

interest group, as is the case for ETX and Q7, 

among others. Here, too, the objective is to 

create common platforms that are interchangeable 

among different manufacturers. In addition, 

some manufacturers offer their own inhouse 

standards which are supported by that 

manufacturer alone, and which are designed 

to allow products by this manufacturer to be 

interchangeable. Among others, E2Brain by 

Kontron or phyCORE and phyCARD by Phytec 

can be found here. In the x86-sector, the supported 

and required interfaces are actually 

driven by one manufacturer and one technology 

and are thereby clearly and actually firmly 

defined. But here too, as mentioned above, a 

standard is often running on empty because 

new interfaces or functions are offered. As a 

rule, this leads to a revision of the standard or 

to a new standard. This can be seen and reproduced 

for the introduction of the PCI bus, as 

a result of which the PC/104 standard was 

first converted into the PC-104 plus and later

into the PCI-104 standard. With the introduction 

of the PCIe bus, XTX was introduced on 

the market as a solution for ETX as well. The 

COM Express standard, which was created in 

long discussions, is also being revised again already 

to make the new interfaces offered by 

Intel available. 

Thus if it is already difficult in the rather clear 

x86 market to define a standard, it is easy to 

see that it is even more difficult in the nonx86 

sector. In many areas, the interfaces offered 

and required here are similar to the x86 interfaces 

and functions, but they go far beyond 

that scope in detail. Thus, in addition to direct 

machine control functions such as PWM, analog 

and digital I/Os and CapCom, field buses 

such as CAN or Profibus are implemented as 

well. But even functions such as keypad used 

in the direct activation of keypads and a multitude 

of graphics and video links are offered 

here. Furthermore, the processors addressing 

the telecom sector offer a multitude of fast 

Ethernet interfaces that are not supported to 

this extent by the known x86 standards. Here, 

it quickly becomes clear that there can be no 

standard in the sense of the x86 standards outside 

the x86 sector, unless one agrees on a 

standard with several hundred pins, which 

makes no sense technologically. 

In spite of the difficulties described, several 

manufacturers offer so-called standards with 

ARM or other non-x86 processors. Thus, the 

first so-called Q7-modules were introduced 

to the market that, however, do not realize all 

interfaces specified in Q7 because there are 

several interfaces missing for the processor 

used. On the contrary, additional interfaces 

not defined in the Q7 standard are routed out 

through the plug. In this case, it is actually 

Figure 2. Embedded module TQM5200S from TQ 


only the mechanical format that is identical to 

Q7, there is little to do with Q7 otherwise. 

This merely results in confusing the market, 

and the systems offered can be realized much 

more optimally in a proprietary version. Further 

so-called standards, such as DIMM systems, 

PhyCard or PhyCore, are attempting to 

tread the PC path. Upon closer inspection of 

the so-called in-house standards offered, one 

will quickly find out that the definitions follow 

the x86 standards and functions closely. Thus, 

as a rule, in addition to the PCI bus, being the 

load-bearing bus, there are one or two Ethernet 

interfaces, two USBs, LVDS, SPI, UART and 

audio. This can be quite sufficient in applications 

that are close to x86. But why do chip 

manufacturers offer a multitude of different 

derivatives and performance categories? As a 

rule, the distinction is in specific interfaces 

and functions that are implemented specifically 

for the market addressed. However, the inhouse 

standards usually do not route out these 

additional functions and can therefore also 

not be used in the applications. This is a clear 

disadvantage since the pin definition for an 

in-house standard can only go through the 

smallest common denominator. 

In addition, if one considers user benefit with 

respect to interchangeability of different families 

of processors and architectures, it quickly 

becomes clear that primarily marketing arguments 

are advanced here. A processor is specifically 

selected according to the application 

which specifies the performance category and 

range of functions required. This will turn out 

differently for different devices, a fact that will 

entail an application-specific base board in 

any event. Ultimately, interchangeability is not 

required here at all, or does not provide any 

advantages. In addition, the different systems 

31 September 2010


require custom system drivers, possibly even 

different operating systems. The benefit is 

therefore not truly visible. Proprietary systems 

such as those offered by TQ, among others, 

follow a different path. Here, the processor 

with its full range of functions occupies center 

stage. A proprietary system elevates the processor 

to a higher system level and realizes everything 

that is required all around the processor, 

regardless of the application on the module. 

All free functions are made available in the application 

via the system connector. 

Thereby, the number of pins and the frame 

size are optimized for the respective processor. 

It makes no functional difference whether the 

processor is soldered directly in the application, 

or whether a module is plugged in. The system 

will boot up automatically and is operational 

n BVM: Mini-ITX motherboard supports 

Intel i7/i5/i3 processors 

BVM’s LV-67C mini-ITX motherboard is based 

on the Intel Q57 and supports versions of the 

Intel i7 high end, the i5 mainstream and the i3 

entry-level 64-bit processors in the LGA 1156 

socket. The 32nm versions of the i5 and i3 

have HD graphics capability with VGA and 

DVI output interface, the 45nm i7 and i5 

versions do not. 

News ID 10833 

n EVOC: Embedded EBX SBC supports 

dual-core i series 

EVOC launched EC5-1813L2NA, an EBX single 

board computer providing high performance 

on low TDP and supporting Intel PGA988 

package 32nm dual-core i series Arrandale CPU: 

Intel Core i7 processor i7-620M and Intel Core 

i5 processor i5-520M. EC5-1813L2NA is 

equipped with two 204-pin DDR3 slots 

(non-ECC), supporting up to 8GB memory. 

News ID 10886 

n IBASE: Atom Mini-ITX motherboard 

with PCIe x4 slot 

IBASE announces the MI889 Mini-ITX motherboard 

that supports a PCI Express x4 that 

can provide higher I/O bandwidth for SATA 

controller cards in RAID server applications. 

The 170 x 170mm MI889 comes with Intel 

Atom N450 (single core) or D510 (dual core) 

processor with Intel 82801HM I/O controller. 

It supports up to 2GB of non-ECC system 

memory in a single DDR2 socket, dual display 

(VGA and LVDS), dual Gigabit Ethernet, four 

serial ports, two SATA interfaces and eight 

USB 2.0 ports. 

News ID 10949 

after applying the single voltage supply voltage. 

Thus, no additional functions have to be realized 

in the application that would serve to 

make the module come to life. Proprietary systems 

can be optimally size-customized for the 

processor used, since after all they do not have 

to take preset standards into account. 

Proprietary systems can be interchangeable 

within a product family such as e.g. for 

Freescale i.MX3xx or PowerQUICC I. This ensures 

that continued use of the module and of 

the base board is possible in case the module 

is exchanged within a family, which can certainly 

happen when individual derivatives are 

discontinued. If it seems useful or if additional 

functions are generally requested on the market, 

proprietary systems can simply be adjusted accordingly. 

This can be done in an internally 

Product News 

n IPC2U: mini-ITX board with AMD CPU 

and SATAII with RAID support 

IPC2U presents the new Mini-ITX KINO-780EB 

board with AMD CPU. The board is based on 

the AMD RS780E + SB710 chipset. It supports 

the AMD Dual Core L325 1.5GHz ASB1 or AMD 

Single Core 210U 1.5GHz ASB 1 CPU. These 

CPU were designed for embedded systems and 

are equipped with the Error Correcting Code. 

News ID 10834 

n Data Modul: Mini-ITX industrial 

motherboard with AMD RS780E chipset 

Data Modul presents the EMX-780E Mini- 

ITX motherboard supporting several single 

core, dual-core and quad-core processors including 

the latest low power AM3 processors, 

such as Phenom II XLT Q54L (65W), Athlon 

II XL V66C (45W), V64L (45W), V50L (25W). 

Current high-performance, multi-core, but 

lower than 65 Watts processors of AMD, like 

dual-core Athlon IIx2 B24 3GHz and quadcore 

Phenom II x4 905e 2.5GHz processors 

are also supported. 

News ID 10881 

n MSC: Qseven modules with 2GB 

DDR2-533 SDRAM 

MSC consistently expands its product portfolio 

of compact embedded Qseven modules in the 

70 x 70mm format based on the Intel Atom 

processor. For memory-demanding applications 

the Qseven module MSC Q7-US15W-FD is now 

not only available with 512 MB or 1 GB memory, 

but also with an onboard 2GB DDR2-533 

SDRAM. Furthermore the Qseven module MSC 

Q7-US15W-FD with integrated 4 GB flash will 

be complemented by a 2 GB flash disk version. 

News ID 10950 


modular fashion, such as for the TQ module 

TQM5200 with freescale processor MPC5200B, 

which allows even greater flexibility and optimization. 

The core will always remain identical, 

regardless of the additional function - in this 

example, the additional graphics software - 

the additional function is realized via the additional 

board surface and additional plugs. 

Furthermore, the additional space required 

provides free space for another memory bank. 

Thus, proprietary systems are useful in the 

non-x86 market since they provide the full 

functionality and strengths of the chosen 

processor on the one hand and, to a certain 

extent, offer the advantages of the interchangeability 

of x86 systems. TQ modules offer optimum 

performance of non-x86 systems on a 

minimum of surface area. n 

n Connect Tech: PC/104 board with 

Qseven module features 

Connect Tech and congatec announce that 

Connect Tech has based the design of its latest 

PCI/104-Express SBC on congatec’s Qseven 

modules. Through their joint efforts, Connect 

Tech added Qseven module features such as 

PCI Express, USB2.0, serial, PS/2 keyboard and 

mouse, Gigabit Ethernet, VGA and LVDS to its 

PCI/104-Express Single Board Computer. 

News ID 11066 

n Advantech: Mini-ITX board with extended 

HD graphics performance 

Advantech introduces a new industrial-grade, 

Mini-ITX motherboard supporting the latest 

Intel Core i7/i5/Celeron mobile processors with 

FCPGA 988 sockets. Intel Core i7, Core i5 and 

Intel Celeron mobile series processors feature 

intelligent performance, power efficiency, and 

integrated Intel HD Graphics with DX10 support. 

AIMB-270 is capable of SATA RAID 0, 1, 

5 & 10 to ensure reliable storage and system 

protection for network-intensive applications. 

News ID 11098 

n Acrosser: PCI/104 solution bundled 

with free Linux package 

Acrosser Technology announces its PC-104 CPU 

module AR-B8020 entry-level platform with 

high cost performance ratio for industrial applications 

that only need 386 or 486 level CPU 

to execute basic industrial applications.. This 

PC-104 CPU board with X86 software structure 

offers a high reliability CPU (RDC R8610 supporting 

133MHz), 64MB onboard SDRAM 

memory, PC/104 and PCI-104 expansion slots, 

fanless design for rugged environment. 

News ID 11053

Adding software security using 

virtualized solutions 

By Stuart Fisher, LynuxWorks 

Security in software systems is 

becoming mandatory. Project 

timescales and code reuse 

policies dictate that large 

amounts of legacy code are 

reused. The challenge of 

integrating this legacy 

software with modern 

military-grade security 

requirements may be 

solved with the latest 

virtualization technology. 

n Traditional software that is non-networked 

intrinsically contains a level of security from 

the threats of malicious code. There is a physical 

boundary that exists between it and other software 

elements within the overall system. It is 

this barrier that provides security from malicious 

code finding its way into the legacy application. 

Sometimes even physical barriers 

are being put in place to prevent user access to 

computer systems, or the software is so highly 

embedded in a product that user access has 

not been an issue. The challenge faced by software 

designers nowadays is how to maintain 

this level of security and safety in a system 

whilst reusing the code in modern designs 

where the system software is all on the same 

processor. 

Over the last number of years we have seen 

processor technology improve by a staggering 

amount. Single CPUs with 100s of times more 

performance than previous designs combined 

with multi-core giving four or even eight processing 

cores on a single chip. This has led 

hardware system designers to consider migrating 

existing bespoke hardware elements onto 

a single box capable of providing the same 

performance as the existing multi-platform 

designs. This gives a significant cost saving 

but more importantly, for example, in many 

aerospace and medical systems it makes it possible 

to add far more electronics into the 

cockpit or patient-monitoring systems for the 

same physical footprint. The challenge of this 

for software designers is making the software 

come together on these single platforms. Where 

code is designed from scratch it is easy to see 

how this can be accomplished, but where 

legacy code is being considered for reuse it is a 

significant issue. One solution to this problem 

is by utilizing advances in software virtualization 

techniques, and in particular we will consider 

the use of a separation kernel. 

Software virtualization is nothing new and 

has long been understood as a way of running 

multiple software products on desktop computers. 

In recent years we have seen virtualization 

migrate into embedded applications and 

start to influence markets such as automotive, 

medical and industrial as well as the more traditional 

aerospace and defense markets. Using 

automotive as an example it can be easily understood 

how reducing the number of processors 

in a vehicle is attractive to car designers, 

but the challenge of now hosting the braking 

systems software and the seat controls on the 

same CPU is clearly an area of concern. 

A separation kernel is very different product 

from a standard virtualization tool that is so 

freely available off the web. There are many 

hypervisor solutions on the market, but all 

these lack the focus on security and in particu- 

SOFTWARE DEVELOPMENT 

Figure 1. Security enforced by 

physical separation 

lar separation of the guest operating systems 

which are being hosted. A common misconception 

in the software world is that virtualization 

implies separation, and that just because 

a platform utilizes virtualization then its software 

subjects must be separated. In the security 

world it is well understood that many virtualization 

architectures and products on the market 

today cannot guarantee any level of software 

separation, and therefore are not candidates 

for systems requiring any level of security or 

safety certification. The separation kernel hypervisor, 

such as LynxSecure from LynuxWorks, 

not only allows multiple guest operating systems 

to run on the same hardware platform, 

but guarantees to the highest level of robustness 

that those guests are also separated and cannot 

affect each other’s functions. 

Not only does a separation kernel separate 

the guest operating systems, it also separates 

the physical devices and information flow 

between the various guests. A software designer 

has the ability to dictate which operating 

system has visibility of certain board 

devices, and which guest operating systems 

are allowed to communicate with each other. 

It is the instantiation of such communication 

paths which facilitate inter-partition communication 

between guests. With such a 

path, the guest would have no visibility or 

knowledge of its peer existence. 

33 September 2010


Figure 2. Example of using a separation kernel 

Using the separation kernel as a base technology, 

the software designer can now guarantee 

that one operating system cannot affect another 

or access certain board devices. Figure 2 illustrates 

a potential architecture to address the 

problem under consideration. In this diagram, 

the Windows subject is running legacy application 

code in a contained Windows environment. 

The OS has no knowledge that it is running 

on a separation kernel or that another 

operating system is running on another core 

on the very same processor. The second operating 

system is designed to be the secure gateway 

and employs complex security software 

to protect the system from the outside world. 

Any data coming from the public network is 

first analyzed by the secure partition, and only 

n emtrion: single board computer now 

with QNX OS 

emtrion now offers QNX support for its single 

board computer, which uses a Renesas SuperH 

7780 processor. In addition to board support 

packages for the operating systems Windows 

CE 6.0 and Linux 2.6, a BSP for QNX Version 

6.4 is also available. The graphic interface is 

based on Photon. As well as supporting USB, 

Ethernet, CAN and Touch, a flash file system 

is also supported on the board. 

News ID 11081 

if it is deemed secure does it make its way via 

inter-partition communication to the Windows 

partition. Using this approach the software designer 

has the flexibility to design the secure 

partition from modern software principles 

whilst the legacy Windows OS is completely 

unchanged. The Windows OS would simply 

see the inter-partition communication path 

as a connection to the outside network and 

has no knowledge that an intermediate software 

guard was analyzing the data and adding a 

level of software security to the not-secure 

legacy software. This premise could indeed be 

extended to any number of theoretical guest 

operating systems each performing a dedicated 

role in the overall system. Some of these guests 

may be legacy code whilst others may be newly 

Product News 

n Wind River: certified VxWorks platform 

mitigates Cyber security risk 

Wind River announces that VxWorks has been 

certified under Wurldtech’s Achilles certification 

program, an internationally recognized 

standard for industrial cyber security. This 

will enable Wind River’s customers in the 

process automation, power and energy, oil 

and gas, transportation, and medical market 

segments to deploy VxWorks with certified 

defenses against cyber attacks. 

News ID 10850 


developed. Systems in the field today already 

employ such technologies in modified designs. 

Products such as a secure separation kernel 

hypervisor now provide a commercial-off-theshelf 

solution to the market not only to reduce 

project time constraints, but also to bring military 

grade software technologies to the commercial 

market. 

We must not forget the role of the hardware 

platform in such designs. Intel VTx and VTd 

architectures give separation kernels the foundation 

to provide the separation between 

guests. Without that hardware assistance it 

would not be possible to protect one guest 

from the other. Take the example of a direct 

memory access (DMA) programmed by malicious 

code within the Windows subject. The 

separation kernel cannot prevent this access 

from taking place as it is performed by the 

hardware, so without hardware assistance any 

certification of such a system would not be 

achievable. However with a hardware I/O memory 

management unit (IOMMU) or in Intel 

technology, VTd, it is possible to prevent such 

accesses. In such a case the IOMMU would 

make sure that any programmed DMA will 

stay within the Windows container. Separation 

kernels must support such technology if they 

are to provide true separation. It is quite evident 

to anybody who looks into the BIOS of most 

motherboards that in almost all cases VTd is 

disabled by default, which would suggest that 

many systems out on the market today simply 

do not offer the type of protection offered by 

leading hardware/software technologies 

In conclusion, virtualization can help not only 

to run multiple productivity tools on desktops, 

but can also address key challenges in the embedded 

software space. Using a product like 

LynxSecure, software designers can address 

the problems of code reuse, hardware consolidation 

and security requirements in new designs, 

whilst maintaining the same architectural 

and runtime environments for those legacy 

applications. n 

n NI: Windows support and Camera Link 

connectivity for Embedded Vision Systems 

National Instruments announces the addition 

of Windows OS support and Camera Link 

connectivity for NI Embedded Vision Systems, 

providing three new options to develop highperformance 

machine vision solutions. Two 

of the new options featuring Windows support, 

the NI EVS-1463 and EVS-1464 systems, offer 

80 GB storage and are suitable for image and 

data logging. 

News ID 10892


through unit testing 

By Mark Pitchford, LDRA 

Unit test tools have long 

provided commercial benefit 

for the team developing the 

highest integrity applications. 

Now these tools can also 

streamline the efforts of their 

peers working in less critical 

environments – even those 

charged with the 

ongoing development of 

undocumented legacy code. 

n Unit test has been around almost as long as 

software development itself. It just makes sense 

to take each application building block, build 

it in isolation, and execute it with test data to 

make sure that it does just what it should do 

without any confusing input from the remainder 

of the application. In the past, the sting 

came from not being able to simply lift a software 

unit from its development environment, 

compile and run it, let alone supply it with 

test data. For that to happen, you need a 

harness program acting as a holding mechanism 

that calls the unit, details any included 

files, writes stubs to handle any procedure 

calls by the unit, and offers any initialisation 

sequences which prepare data structures for 

the unit under test to act upon. Not only was 

creating that process laborious, but it took a 

lot of skill. More often than not, the harness 

program required at least as much testing as 

the unit under test. Perhaps more importantly, 

a fundamental requirement of software testing 

is to provide an objective, independent view of 

the software. The very intimate code knowledge 

required to manually construct a harness compromised 

the independence of the test process, 

undermining the legitimacy of the exercise. 

In developing applications for the medical, 

railway, aerospace and defence industries, unit 

test is a mandatory part of a software development 

cycle - a necessary evil. For these high 

integrity systems, unit test is compulsory, and 

the only question is how it might be completed 

in the most efficient manner possible. It is 

therefore no coincidence that many of the 

companies developing tools to provide such 

efficiency have grown from this niche market. 

In non-safety-critical environments, perceived 

wisdom is that unit testing is a nice idea in 

principle, but commercially unjustifiable. A 

significant factor in that stance is the natural 

optimism which abounds at the beginning of 

any project. At that stage, why would anyone 

spend money on careful unit testing? There 

are great engineers in the team, the design is 

solid, sound management is in place. What 

could possibly go wrong? However, things can, 

and do go wrong, and while unit test cannot 

guarantee success, it certainly helps minimise 

failure. So if we look at the tools designed and 

proven to provide quick and easy unit tests in 

high integrity systems, it makes sense that the 

same unit tests would provide a solid solution 

for those working on commercially developed 

software as well. 

Unit testing cannot always be justified. And 

sometimes it remains possible to perform unit 

test from first principles, without the aid of 

any test tool at all. There are pragmatic judgements 

to be made. Sometimes that judgment is 

easy. If the software fails, what are the implications? 

Will anyone be killed, as might be the 


Figure 1. A single test case 

(inset) can exercise some or all 

of the call chain associated 

with it. In this example, 

AdjustLighting, note the red 

colouring highlights exercised 

code. 

case in aircraft flight control? Will the commercial 

implications be disproportionately high, as 

exemplified by a continuous plastics production 

plant? Or are the costs of recall extremely high, 

perhaps in a car engine controller? In these 

cases, extensive unit testing is essential and any 

tools that aid in that purpose make sense. 

On the other hand, if software is developed 

purely for internal use or is perhaps a prototype, 

then the overhead in unit testing all but the 

most vital of procedures would be prohibitive. 

As you might expect, there is a grey area. Suppose 

the application software controls a mechanical 

measuring machine where the quantity 

of the devices sold is low and the area served 

is localised. The question becomes: would the 

occasional failure be more acceptable than the 

overhead of unit test? In these circumstances, 

it’s useful to prioritise the parts of the software 

which are either critical or complex. If a software 

error leads to a strangely coloured display 

or a need for an occasional reboot, it may be 

inconvenient but not in itself justification for 

unit test. On the other hand, the unit test of 

code which generates reports showing whether 

machined components are within tolerance 

may be vital. Again, it comes down to cost. 

The later a defect is found in the product development, 

the more costly it is to fix - a concept 

first established in 1975 with the publication 

of Brooks Mythical Man Month and 

35 September 2010


Figure 2. The later a defect is identified, the higher the cost of rectifying it. 

Figure 3. Unit test tools lend themselves admirably to test-driven development by providing a 

mechanism to write test cases before any source code is available. 

proven many times since through various studies. 

The automation of any process changes 

the dynamic of commercial justification. This 

is especially true of test tools since they make 

earlier unit test much more feasible. 

Consequently, modern unit test almost implies 

the use of such a tool unless only a handful of 

procedures are involved. The primary function 

of such unit test tools is to automatically generate 

the harness code which provides the 

main and associated calling functions or procedures 

(generical procedures). These facilitate 

compilation and allow unit testing to take 

place. The tools not only provide the harness 

itself, but also statically analyse the source 

code to provide the details of each input and 

output parameter or global variable in any 

easily understood form. Where unit testing is 

performed on an isolated snippet of code, 

stubbing of called procedures can be an im- 

portant aspect of unit testing. This can also be 

automated to further enhance the efficiency 

of the approach. This automation makes the 

assignment of values to the procedure under 

test a simple process, and one which demands 

little knowledge of the code on the part of the 

test tool operator. This creates that necessary 

unit test objectivity because it divorces the 

test process from that of code development 

where circumstances require it, and from a 

pragmatic perspective substantially lowers the 

level of skill required to develop unit tests. It is 

this ease of use which means that unit test can 

now be considered viable for development 

since each procedure can be tested at the time 

of writing. When these early unit tests identify 

weak code, it can be corrected whilst the original 

intent remains very fresh in the mind of the 

developer. For some, the terms unit test and 

module test are synonymous. For others, the 

term unit implies the testing of a single proce- 


dure, whereas module suggests a collection of 

related procedures, perhaps designed to perform 

some particular purpose within the application. 

Using the latter definitions, manually 

developed module tests are likely to be easier 

to construct than unit tests, especially if the 

module represents a functional aspect of the 

application itself. In this case, most of the calls 

to procedures are related and the code accesses 

related data structures which makes the preparation 

of the harness code more straightforward. 

Test tools render the distinction between unit 

and module tests redundant. It is perfectly possible 

to test a single procedure in isolation and 

equally possible to use the exact same processes 

to test multiple procedures, a file or multiple 

files of procedures, a class (where appropriate), 

or a functional subset of an entire system. As a 

result, the distinction between unit and module 

test is one which has become increasingly irrelevant 

to the extent that the term unit test has 

come to include both concepts. 

Such flexibility facilitates progressive integration 

testing. Procedures are first unit tested and 

then collated as part of the subsystems, which 

in turn are brought together to perform system 

tests. It also provides options when a pragmatic 

approach is required for less critical applications. 

A single set of test cases can exercise a 

specified procedure, all procedures called as a 

result of exercising the single procedure, or 

anything in between. The use of test cases 

which prove the functionality of the whole 

call chain are easily constructed. Again, it is 

easy to mix and match the processes depending 

on the criticality of the code under review. 

This all embracing unit test approach can be 

extended to multithreaded applications. In a 

single-threaded application, the execution path 

is well-defined and sequential, such that no 

part of the code may be executed concurrently 

with any other part. In applications with multiple 

threads, there may be two or more paths 

executed concurrently, with interaction between 

the threads a commonplace feature of the system. 

Unit test in this environment can ensure 

that particular procedures behave in an appropriate 

manner both internally and in terms of 

their interaction with other threads. 

Sometimes, testing a procedure in isolation is 

impractical. For instance, if a particular procedure 

relies on the existence of some ordered 

data before it can perform its task, then similar 

data must be in place for any unit test of that 

procedure to be meaningful. Just as unit test 

tools can encompass many different procedures 

as part of a single test, they can also use a sequence 

of tests with each one having an effect 

on the environment for those executed subsequently. 

For example, unit testing a procedure 

which accesses a data structure may be achieved 

by first implementing a test case to call an ini-

tialisation procedure within the application, 

and then a second test case to exercise the procedure 

of interest. Unit test does not imply 

testing in only the development environment. 

Integration between test tools and development 

environments means that unit testing of software 

can take place seamlessly using the compiler 

and target hardware. 

Whilst unit testing at the time of development 

is a sound principle to follow, all too often ongoing 

development compromises the functionality 

of software which is considered complete. 

Such problems are particularly prevalent when 

adding functionality to code originally written 

with no knowledge of later enhancements. Regression 

testing is what is needed here. By 

using a test case file to store a sequence of 

tests created for the original SOUP software 

(software of unknown pedigree), it is possible 

to recall and reapply it to the revised code to 

prove that none of the original functionality 

has been compromised. Once configured, this 

regression testing can be initiated as a background 

task and run perhaps every evening. 

Reports can highlight any changes to the 

output generated by earlier test runs. In this 

way, any code modifications leading to unintentional 

changes in application behaviour 

can be identified and rectified immediately. 

Modern unit test tools come equipped with 

user friendly, point-and-click graphical user 

interfaces, which are easy and intuitive to use. 

However, if faced with thousands of test cases, 

a GUI interface is not always the most efficient 

way to handle the development of test cases. 

In recognition of this, test tools are designed 

to allow these test case files to be directly developed 

from applications such as Microsoft 

Excel. As before, the regression test mechanism 

can then be used to run the test cases held in 

these files. 

In addition to using unit test tools to prove developed 

code, they can also be used to develop 

test cases for code still in conception phase - 

an approach known as test-driven development 

(TDD). TDD is a software development technique 

that uses short development iterations 

based on pre-written unit test cases that define 

desired improvements or new functions. Each 

iteration produces code necessary to pass that 

iteration tests. The programmer or team refactors 

the code to accommodate changes. Traditionally, 

many applications have been tested 

by functional means only. The source code is 

written in accordance with the specification, 

and then tested to see if it all works. The problem 

with this approach is that no matter how 

carefully the test data is chosen, the percentage 

of code actually exercised can be very limited. 

That issue is compounded by the fact that the 

procedures tested in this way are only likely to 

handle data within the range of the current 

application and test environment. If anything 

changes a little - perhaps in the way the application 

is used, or perhaps as a result of slight 

modifications to the code - and the application 

could be running to an entirely untested execution 

in the field. Of course, if all parts of the 

system are unit tested and collated on a piecemeal 

basis through integration testing, then 

this will not happen. But what if timescales 

and resources do not permit such an exercise? 

Unit test tools often provide the facility to instrument 

code. This instrumented code is 

equipped to track execution paths, providing 

evidence of the parts of the application which 

have been exercised during execution. Such 

an approach provides the information to produce 

such data. Code coverage is an important 

part of the testing process in that it shows the 

percentage of the code that has been exercised 

and proven during test. Proof that all code has 

been exercised correctly need not be based on 

unit tests alone. To that end, some unit tests 

can be used in combination with system test 

to provide a required level of execution coverage 

for a system as a whole. This means that 

the system testing of an application can be 

complemented by unit tests to exercise code 

which would not normally be exercised in the 

running of the application. Examples include 

defensive code (e.g. to prevent crashes due to 

inadvertent division by zero), exception handlers 

and interrupt handlers. 

Generally, the output data generated through 

unit tests is an important end in itself, but this 

is not necessarily always the case. There may 

be occasions when the fact that the unit tests 

have successfully completed is more important 

than the test data itself. This happens when 

source code is to be tested for robustness. To 

provide for such eventualities, it is possible to 

use test tools to automatically generate test 

data as well as the test cases. High levels of 

code execution coverage can be achieved by 

this means alone, and the resultant test cases 

can be complemented by means of manually 

generated test cases in the usual way. An interesting 

application for this technology involves 

legacy code. Such code is often a valuable 

asset, proven in the field over many years but 

wwww.embedded-control-europe-com w w. 

embedded 

-con 

n trol-eur 

ope-com 


likely to have been developed on an experimental, 

ad hoc basis by a series of expert 

gurus – expert at getting things done and in 

the application itself, but not necessarily at 

complying with modern development practices. 

Frequently this software of unknown pedigree 

(SOUP) is required to form the basis of new 

developments which are obliged to meet modern 

standards either due to client demands or 

because of a policy of continuous improvement 

within the developer organisation. This situation 

may be further exacerbated by the fact 

that coding standards themselves are the subject 

of ongoing evolution, as the advent of MISRA 

C:2004 clearly demonstrates. 

If there is a need to redevelop code to meet 

such standards, then this is a need to not only 

identify the aspects of the code which do not 

meet them, but also to ensure that in doing so 

the functionality of the software is not altered 

in unintended ways. The existing code may 

well be the soundest or only documentation 

available and so a means needs to be provided 

to ensure that it is dealt with as such. Automatically 

generated test cases can be used to 

address just such an eventuality. By generating 

test cases using the legacy code and applying 

them to the rewritten version, it can be proven 

that the only changes in functionality are those 

deemed desirable at the outset. 

The Apollo missions may have seemed irrelevant 

at the time, and yet hundreds of everyday 

products were developed or modified using 

aerospace research - from baby formula to 

swimsuits. Formula One racing is considered 

a rich man’s playground, and yet British soldiers 

are benefiting from the protective qualities of 

the light yet strong materials first developed 

for racing cars. Hospital patients and premature 

babies may stand a better chance of survival 

than they would have done a few years ago, 

thanks to the transfer of F1 know-how to the 

medical world. 

Likewise, unit test has long been perceived to 

be a worthy ideal - an exercise for those few involved 

with the development of high-integrity 

applications with budgets to match. But the advent 

of unit test tools means that the latest unit 

test tools provide slick, efficient mechanisms 

that optimise the development process for all. 

The availability of such tools has made this 

technology and unit testing itself an attractive 

proposition for applications where sound, reliable 

code is a commercial requirement, rather 

than a life-and-death imperative. n 

37 September 2010

ADVANCEDTCA 

Performance gains from multi-core 

processors partnered with ATCA 

By Gene Juknevicius, GE Intelligent Platforms 

Today, multi-core processors 

are an integral element of 

electronics design and are 

well supported by the 

AdvancedTCA infrastructure. 

This article describes technologies 

and tools designers 

can use to get full advantage 

from multi-core processors 

combined with ATCA. 

n Multi-core processor technology combined 

with the AdvancedTCA form factor results in 

multi-faceted performance scaling options: 

performance can be scaled by using processor 

silicon with more processor cores, as well as 

by adding more ATCA blades into the chassis. 

Moreover, ATCA systems are easy to configure 

for a specific workload by combining standard 

multi-core x86 processors with specialized 

packet processors. Having multiple cores within 

a processor is potentially highly advantageous, 

of course, but they are useless unless the software 

infrastructure has a means of utilizing 

these cores. Virtualization is one technique 

that allows use of multiple cores to run multiple 

applications and their operating systems in 

parallel. New application development - or 

porting an existing application to a multi-core 

environment - is eased by the development 

tools that are available. Packet processors in 

particular have a powerful set of tools that 

allow the design of applications that run in 

parallel on multiple cores. 

Just a few years ago each new processor silicon 

release brought along a worthwhile clock frequency 

improvement. Today, however, clock 

frequency is not the main news in a new generation 

processor release; it is the number of 

processor cores within the device that is taking 

center stage. As usual, small startups such as 

Cavium Networks and RMI (now NetLogic) 

were the first to market with multi-core general 

purpose processors. Then followed the giants: 

Intel, AMD and Freescale. Today, 4-8 cores 

within a processor are the norm - and there 

are architectures available that feature as many 

as 64 cores within one processor. The motivation 

for multi-core processors is fairly simple: 

when running a typical application, the processor 

spends most of its time waiting for data to 

process. Historically, memory latency has improved 

at a much slower pace than the speed 

of the processor. Today, the mismatch between 

processor and memory is such that adding a 

few extra clocks to the processor does not improve 

performance to any worthwhile degree. 

As if this is not a big enough problem, there is 

the issue of power consumption: adding a few 

extra hertz to the clock translates into a significant 

increase in power consumption. 

From the multi-core architecture perspective, 

having multiple cores, each running perhaps 

at a slightly slower speed, results in a higher 

overall performance solution. Considering that 

the processor spends roughly three-quarters 

of its time waiting for the memory, this approach 

works well for applications that can 

benefit from parallel processing. Obviously, 

the memory subsystem implementation has 

to support multiple data accesses in parallel, 

which is typically the case today. Let us move 

the focus from the silicon to the system. When 


Figure 1. ATCA integrated platform from NEI 

a single server with two or four multi-core 

processors is required, the 19-inch rack-mountable 

enclosure – the pizza box - works very 

well. When the application requires more than 

that, or when redundancy and higher reliability 

are required, AdvancedTCA becomes a good 

choice for system implementation. The AdvancedTCA 

chassis can support up to 14 dualprocessor 

blades interconnected via two highperformance 

Ethernet switches in a redundant 

fashion. All blades within the chassis share 

power supplies and cooling fans, which are 

also implemented to support redundancy and 

higher reliability. 

A key requirement when building a multiblade 

system is a high-speed, reliable interconnect 

between the blades. From this perspective, 

an ATCA system interconnects each blade via 

a fabric interface and base interface. The fabric 

interface, which is considered to be a data 

path interface, is predominantly 10Gbit Ethernet 

today and will soon support 40Gbit Ethernet. 

The base interface is a control path and is 

implemented using 1Gbit Ethernet. Both fabric 

and base interfaces are implemented in a redundant 

fashion, such that each ATCA blade 

connects to both ATCA hubs which provide 

the required Ethernet switching resources. All 

connectivity is provided via the ATCA backplane, 

reducing external cabling, thereby making 

the overall system more reliable and more

serviceable. The separation of the control plane 

and data plane not only enables high performance 

blade management and control services, 

but also isolates the control traffic from the 

revenue-generating data plane traffic. Such 

isolation of the two planes becomes critical 

when overall system security is considered. 

Plane isolation ensures that data plane traffic, 

which is typically customer-facing traffic, will 

not intentionally or unintentionally start 

managing Ethernet switches and disrupt the 

operation of the complete system. 

Depending on the application type, the high 

performance interconnect brings a different 

value proposition. In a compute type application, 

it is essential that large numbers of processors 

communicate with high throughput and 

very low latency. To that extent, 10Gbit and 

40Gbit Ethernet can provide the required data 

throughput. Some Ethernet switches also support 

pass-through switching mode where packet 

transmission starts before the packet is fully 

received. In such cases, packet switching latency 

can be lower than 500ns. Although configuring 

two hubs (Ethernet switches) in an ATCA 

chassis is primarily for redundancy, it is also 

possible to use both hubs in parallel, effectively 

doubling the available bandwidth. From the 

compute power density perspective, it is interesting 

to note that 14 GE Intelligent Platforms 

A10200 ATCA blades, each featuring dual Intel 

6-core Westmere processors, yield no fewer 

than 168 x86 cores within a single ATCA 

chassis, all interconnected via an in-chassis 

high-speed interconnect. 

Compute applications also tend to require 

significant storage capacity, bandwidth and 

reliability. There are three main ways to address 

storage requirements. At the lowest level, each 

ATCA blade can have local hard disks, located 

on the blade itself or on an associated RTM: 

these could be two redundant SAS drives. At 

the next level, one or more storage ATCA 

blades could be used within the system. Such 

storage blades would be accessed via Ethernet 

using either the FCoE or iSCSI protocols. 

ATCA storage blades can be shared among 

multiple processor blades. Finally, an external 

storage array can be connected via Fibre 

Channel, FCoE or iSCSI. 

A key feature of communication applications 

is their requirement for high data throughput 

and packet processing. Also, they typically 

lend themselves well to parallel processing 

which is where multi-core technology finds its 

ADVANCEDTCA 

optimal advantage. Although processors from 

both AMD and Intel are excellent computing 

devices - especially when multiple cores are 

considered - both lack the ability to efficiently 

get data in and out at very high data rates. 

Packet processors, another type of multi-core 

processor architecture, are specifically optimized 

to address the problem of efficient 

movement in and out of packetized data. Such 

devices are readily available in the ATCA blade 

form factor allowing system designers to take 

advantage of both x86 compute resources and 

packet processor packet manipulation resources 

within the same system. The interoperability 

inherent in the ATCA specification enables 

designers to plug in multiple x86 processor 

blades as well as multiple packet processor 

blades and interconnect them via high 

performance Ethernet interfaces. 

From this perspective, Ethernet switches within 

hubs provide additional value in load distribution. 

Ethernet switches today employ sophisticated 

access control list features that 

allow packets, based on their Layer-2 to Layer- 

4 information, to be steered to a specific ATCA 

blade. Such policy-based routing allows packet 

streams to be distributed at very high data 

rates (10Gbit/sec to 100Gbit/sec) among

ADVANCEDTCA 

Figure 2. 16-slot ATCA chassis internal interconnect diagram 

multiple ATCA blades, while ensuring that 

packets belonging to the same flow are always 

directed to the same blade. An example of a 

high-performance communication system is 

shown in figure 4. This ATCA system employs 

two Ethernet hubs, two GE A10200 multicore 

x86 processor blades and up to 12 GE 

AT2-5800 packet processor blades with dual 

16-core OCTEON Plus processors. Simple 

math reveals that this system provides 320 

MIPS64 cores (OCTEON devices) and 24 x86 

cores (Intel Westmere devices). 

From the data processing perspective, data enters 

the system via Ethernet hubs where packets 

are distributed - based on policies - among 

packet processing blades. Then, within the 

packet processor blade, packets are further 

distributed between two OCTEON devices 

and finally, within each OCTEON device, between 

the cores. The packet processors perform 

the majority of the high throughput packet 

processing, and specific packets requiring more 

extensive processing power are forwarded to 

x86-based blades. The key principle here is 

that although the majority of packets require 

little processing, a small subset requires more 

significant processing power. It is clear that 

any ATCA system is useless without software. 

Having hundreds of processor cores offers 

Figure 3. The GE Intelligent Platforms A10200 ATCA single 

board computer 

huge potential, but unless used efficiently they 

are a waste of silicon. Historically, most 

applications were written without any parallel 

computing concepts in mind. Consequently, 

although modern compilers attempt to recognize 

areas in the code that lend themselves to 

parallel processing and try to harness the 

power of multiple cores, performance improvements 

are very limited when running legacy 

applications on multi-core hardware. Virtualization 

is often used today to better utilize 

multiple processor cores. In a virtualized environment, 

multiple instances of the operating 

system – or even multiple dissimilar operating 

systems - run on the same multi-core processor. 

Since each operating system has no relationship 

with the others, the operating systems can be 

happily executed in parallel on multiple cores. 

Hardware, with the help of Hypervisor, ensures 

that each operating system can safely access 

its own memory and I/O devices without 

disturbing its neighbors. Virtualization allows 

the consolidation of multiple physical servers 

into one server with multi-core processors. 

ATCA allows further consolidation of multiple 

blades with multiple multi-core processors: 

racks of legacy servers can be reduced to a 

single ATCA chassis. Virtualization within the 

ATCA environment provides another benefit - 

redundancy and high availability. 

Using a high availability 

virtualized operating 

system, an application can 

be migrated from one physical 

server to the other if 

hardware failure occurs. 


Figure 4. ATCA System with two Westmere blades and ten OCTEON blades 

Since packet processors 

were designed for parallel 

processing from the start, 

their software environment 

and development tools are 

fully geared toward application 

development in a 

multi-core environment. Al- 

though OCTEON and similar devices are often 

called packet processors, internally they are 

based on standard processor architectures, 

such as MIPS64, and can run standard operating 

systems such as Linux. Their performance 

advantages, however, are best exposed when 

running simplified proprietary operating systems, 

such as Cavium Simple Executive. It is 

important not to confuse these devices and 

their operating systems with the network 

processors of the past, such as Intel XScale. 

Modern packet processors are programmed 

using standard C and C++ language even 

when their proprietary operating system is 

being used; in fact, they allow existing C code 

to be simply ported. 

Simplistic applications, such as a packet filter 

or L-2, L-3 switch, can be developed as sequential 

code which runs to completion and 

executes in an endless loop. The same code 

could be run on all cores, and the parallel nature 

of the processing would be provided by 

the hardware itself, which would schedule a 

packet processing event onto the next available 

processor core, enforcing packet ordering and 

atomicity rules if desired. The hardware also 

takes care of memory management and cache 

coherency, allowing developers to focus on 

the application itself. Inter-core communication 

can be implemented by setting aside a shared 

memory region or by using a shared variable. 

Depending on the application and development 

requirements, a number of software packages 

can help developers get a head start. One notable 

example is 6WINDGate software which allows 

the seamless marriage of x86 processors with 

packet processors, offloading time-critical tasks 

to be run by Simple Executive on the packet 

processors and providing a large number of frequently 

needed protocols. 6WINDGate can be 

used standalone, or as a base platform for a specific 

application, and can abstract inter-processor 

and inter-core communications, significantly 

simplifying software development effort. n

n Green Hills and RTI ease integration 

of mission-critical devices 

Green Hills Software and Real-Time Innovations 

have expanded their longstanding partnership. 

As a result, RTI’s DDS-compliant 

middleware will gain optimized support for 

Green Hills Software’s latest release of its safety 

and security-certified INTEGRITY real-time 

operating system. These enhancements include 

advanced support for INTEGRITY RTOS capabilities, 

such as Symmetric Multiprocessing 

and Asymmetric Multiprocessing. RTI will 

also offer standard release support for the IN- 

TEGRITY RTOS on a broad range of microprocessor 

families, including those based on 

Intel Architecture (Core i3, i5, i7, Atom and 

Xeon) and a wide range of Power Architecture 

microprocessors. 

News ID 10874 

n DDC: bus analyzer software support 

for MIL-STD-1553 cards 

Data Device Corporation announces dataSIMS 

Avionics Data Bus Test and Analysis Software 

support for its complete new line of AceXtreme 

data bus cards. DDC’s new line of AceXtreme 

BU-67X Series cards offer advanced MIL- 

STD-1553 functionality, including BC, or 

Multi-RT and concurrent Monitor per channel, 

along with common API for test and embedded 

applications. Now, with the addition of 

dataSIMS support, DDC can offer an all-inone 

test and analysis solution for all your 

data bus needs. 

News ID 11036 

n Wind River: safety critical platform 

for DO-178B and IEC 61508 

Wind River announced VxWorks Cert Platform, 

a combination of Wind River’s existing 

VxWorks DO-178B and IEC 61508 safety critical 

platforms in one product offering. The 

new software platform introduces support for 

C++ and real-time processes providing more 

flexibility in programming and design and 

enabling improved compatibility with existing, 

imported or reused code. 

News ID 11106 

n BVM: 3.5 inch miniboard SBC 

with N450/D410 and D510 Atom 

BVM announces the 3.5 inch Embedded Miniboard 

LE-376, based around the single core 

N450 and D410 and the dual core D510 Intel 

Atom processors, all running at 1.66GHz and 

addressing up to 2GB for the N450 or 4GB for 

the D410 and D510 of DDR2 memory. Low 

power consumption and integrated graphics 

and video processing makes this platform well 

suited for many embedded market segments; 

the N450 Thermal Design Power rating is only 

5.5 Watt, the D410 is 10 Watt and D510 dual 

core is only 13 Watt. 

News ID 10937 

n Advantech: rugged PC/104 CPU board 

with DM&P Vortex86DX processor 

Advantech has released a cost-effective SoC 

PC/104 CPU module driven by the DM&P 

Vortex86DX 1.0 GHz processor with 256 MB 

of DDR2 SDRAM memory on board. PCM- 

3343 has the standard dimensions of 96 x 90 

mm in a fanless PC/104 architecture, supporting 

rich graphic output including VGA and 24-bit 

LVDS or TTL up to 1024 x 768 resolution. 

PCM-3343 delivers ultra low power consumption 

of less than 5 Watts under full load, and 

extended temperature support between -40. 

News ID 10790 

n AAEON: 3.5-inch SubCompact board 

has expansion options 

AAEON launches the 3.5â€ SubCompact 

board GENE-9655, adopting the Intel GME965 

+ ICH8M chipset and the Socket-P that can 

accept an Intel Core 2 Duo/ Celeron M processor, 

such as the 2.2GHz T7500. The board can 

also be ordered with the option to have an 

Intel BGA processor embedded on it in place 

of the processor socket. The GENE-9655 supports 

up to 2 GB of DDR2 533/667 memory 

in its one 200-pin SODIMM socket. 

News ID 10846 

n congatec launches power-efficient 

COM Express modules 

congatec expands its COM Express product 

range with the conga-BS57 module featuring 

the latest Intel low-power processors in a 

small form factor BGA package. Depending 

on the application, the conga-BS57 utilizes a 

range of processors starting with Intel Celeron 

U3400 processor (4MB Cache, 1.06 GHz, TDP 

18 Watt) up to the Intel Core i7-620 LE 

processor (4MB Cache, 2.0 GHz, TDP 25 

Watt). The conga-BS57 features the Mobile 

Intel HM55 Express Chipset which measures 

only 25 x 27 mm and provides a powerful, 

compact two-chip solution with support for 

up to 8GByte (1066 MT/s) dual channel DDR3 

memory 

News ID 10804 

n IEI: Dual DVI display multimedia box 

IEI Technology introduces the ECN-581A- 

QM57, a dual DVI Display Multimedia Box. 

The ECN-581A-QM57 delivers enhanced visual 

performance with its new generation 

Intel Core i5-520M mobile processor and 

mobile IntelQM57 Express Chipset. The Intel 

QM57 chipset is the advanced version of 

Intel HM55 and supports Intel AMT 6.0.The 

ECN-581A-QM57 has a high performance 

graphics engine with full AVC, VC-1 and 

MPEG2 decoding acceleration as well as 1080p 

HD digital output for large-sized displays 

making the ECN-581A-QM57 an ideal system 

for digital signage. 

News ID 10974 

Editors 

Jürgen Hübner 

phone +49(0)8092-2477413 

fax +49(0)8092-2477429 

jh@iccmedia.com 

Wolfgang Patelay 

wp@iccmedia.com 

Tony Devereux 

devrex@teyboyz.freeserve.co.uk 

PRODUCT NEWS 

For Reader Inquiries and Address Changes 

please contact: 

info@iccmedia.com 

Sales & Marketing Director 

Manfred Blumoser 

phone +49(0)8092-2477411 

fax +49(0)8092-2477429 

mb@iccmedia.com 

Claudia Mellein 

cm@iccmedia.com 

Christiane Lockner 

cl@iccmedia.com 

Sales Office - UK and USA, Benelux, 

Scandinavia 

Blue Sky Communications 

Malcolm Cameron 

Kent Innovation Centre 

Thanet Reach Business Park 

Millennium Way 

Broadstairs Kent CT10 2QQ UK 

phone +44 (0)1843 609357 

fax +44 (0)1843 609358 

mc@blue-sky-communications.com 

Sales Office - Asia 

Jean Cheng, jean@i-media.com.tw 

Vivian Hung, vivian@i-media.com.tw 

Innovative Media Information & Services 

7F-3, No. 26, Sec. 2, Ming-Quan East Rd. 

Taipei 104 Taiwan 

phone +886 2 2563 1186 

Head Office 

ICC Media GmbH 

Rauwagnerstr. 5 

85560 Ebersberg / Germany 

Editorial Office UK 

36a Blackacre Road 

Theydon Bois 

Essex, CM16 7LU 

Copyright© All rights reserved. 

No part of this publication may be reproduced or transmitted in 

any form or by any means without the prior express written 

permission of ICC Media. 

Although we make every effort to present up-to- date, accurate 

information, boards&solutions will not be responsible for any 

errors or omissions or for any results obtained from the use of 

such information. The magazine will not be liable for any loss 

caused by the reliance on information obtained on this site. 

Furthermore, boards&solutions does not warrant the accuracy 

or completeness of the information, text, graphics in this magazine. 

The opinions expressed in the articles are those of the 

authors and not necessarily the opinions of the publisher. 

41 September 2010

PRODUCT NEWS 

n IBASE: Core i7-610E processors-based 

Mini-ITX motherboard 

IBASE releases a new Intel Core i7-610E processor-based 

MI957 Mini-ITX motherboard with 

the mobile Intel QM57 Express Chipset. MI957 

is designed for applications requiring highlevel 

processing and graphics performance 

with a long product life such as those in digital 

signage, gaming, POS, kiosk and multimedia 

sectors. 

News ID 11076 

n GE: new OCTEON II based processor 

for network applications 

GE Intelligent Platforms announces the WANic 

6354 intelligent high performance packet 

processor designed for demanding high volume 

IP network applications. Based on the latest 

multi-core Cavium OCTEON II CN6335-AAP 

application accelerator processor, its combination 

of advanced technologies allows for 

higher speed networks and for the inclusion 

of advanced security functionality in those 

networks. 

News ID 10956 

n Bicker: 150W power supply with UPS 

function 

Bicker Elektronik presents the ATX power 

supply PCFL-180P-X2S2 which combines the 

functions of a PC power supply with those of 

a UPS. In UPS mode the electrical energy is 

supplied by a separate NiMH battery pack. 

The PCFL-180P-X2S2 is designed for industrial 

applications with a service life of at least 7 

years in 24 h continuous operation at a temperature 

of +25 °C. The unit is distinguished 

further by a robust design and high-quality 

components: The PCB is made of FR4 material 

and is plated through. 

News ID 10893 

n AAEON: Pentium M/ Celeron M network 

appliance with 4 LAN ports 

AAEON releases the FWS-2300, a desktop network 

security solution with four LAN ports. 

AAEON had previously released a condensed 

desktop network appliance called the FWS- 

2150 for general use. However, The FWS-2300 

is an advanced model that adopts the Intel 

platform to provide a secured network solution 

to customers who require higher needs in network 

security. 

News ID 10831 

n N.A.T. MicroTCA Carrier Hub supports 

Gen2 PCIe and Gen2 SRIO 

N.A.T. announces the 4th generation (Gen4) 

of its MicroTCA Carrier Hub NAT-MCH supporting 

the second generation (Gen2) of PCIe 

and SRIO switches. The NAT-MCH is the 

central switching resource for any MicroTCA 

system and dedicated to any application in 

the communication, aerospace and defence, 

medical or industrial market. 

News ID 11071 

n Moxa: industrial multi-service gateway 

Moxa announces an industrial multi-service 

gateway that is powerful enough to construct 

decentralized SCADA systems for seamless 

communications at field sites and between the 

field site and control center in wide-area 

distributed applications such as oil and gas, 

pipeline, and trackside monitoring. The 4-slot 

VPort 704 interfaces with various communication 

modules to collect and transform all 

serial I/O, video, and other types of data into 

Ethernet packets for fast and reliable 

transmission over IP networks. 

News ID 10799 

n Advantech: ETX CPU module with Atom 

N450/D510 

Advantech introduces its most cost-effective ETX 

CPU module, SOM-4463, supporting the Intel 

Atom N450/D510 single and dual core processors 

with ICH8M chipset. SOM-4463 offers low 

power consumption and is the first Intel Atom 

based dual-core D510 solution. The single core 

N450 processor offers fanless capability. In a 

basic form factor of 114 mm x 95 mm, SOM- 

4463 fits the embedded ETX 3.0 standard. 

News ID 10780 

n Moxa: Embedded device servers without 

RJ45 Ethernet ports 

Moxa’s MiiNePort E2 embedded device servers, 

which do not have an RJ45 Ethernet connector, 

are designed for manufacturers who want to 

add sophisticated network connectivity to their 

serial devices with minimal integration effort. 

The MiiNePort E2 is empowered by the MiiNe, 

Moxa’s second generation SoC, which supports 

10/100 Mbps Ethernet, up to 921.6 Kbps serial 

baudrate, a versatile selection of ready-to-use 

operation modes, and requires only a small 

amount of power. 

News ID 10839 

FREE Subscription to boards & solutions magazine 


Advertisers Index 

COMPANY PAGE 

Advantech 2 

Avalue 43 

congatec 29 

Data Device Corporation 7 

EVOC 9 

Kontron 31 

MEN Mikro Elektronik 25 

Messe München 44 

MSC 3 

N.A.T. 39 

Schroff 21 

VIA 17 

Ensure getting your personal copy of bas magazine free of charge by completting the online form at: 

• www.embedded-control-europe.com/bas-magazine 

www www.embedded-control-europe.com/ba 

p / as-magazine g

n VIA: Nano E-Series processor-based 

embedded system with dual Gigabit LAN 

The VIA EPIA-M840 packs the latest VIA 

Nano E-Series processor, dual Gigabit LAN, 

eight COM ports and two dual channel 24-bit 

LVDS channels making an ideal solution for a 

range of embedded applications including the 

latest industrial automation, kiosk, POI and 

HMI applications. 

News ID 11074 

n Interface Masters: quad segment 10 Gb 

link aggregation TAP 

Interface Masters announces an additional 

member of its 10 Gigabit product family that 

is designed to integrate with Network Monitoring 

appliances, Analytics, probes and logging 

tools, Intrusion Detection Systems and 

Central Office applications. Niagara 3218 consists 

of Quad Link Aggregation TAP segments 

which possess full 10 Gigabit and 1 Gigabit 

monitoring capabilities. 

News ID 10951 

n AAEON: eco-friendly embedded controller 

AAEON has unveiled an eco-friendly solution, 

GES-3300F, that belongs to AAEON’s Green 

Embedded Systems product line. The GES- 

3300F adopts the Intel Core2 duo/ Core Duo/ 

Celeron M processor with two 240-pin DDR2 

400/533/667 DIMM (up to 4GB) system memory. 

It features the Intel 945GME and ICH7M 

chipset to provide high performance and 

consists of two Gigabit Ethernets for faster 

network communications. 

News ID 11045 

n TQ: small PowerQUICC II Pro modules 

with Gigabit Ethernet 

TQ’s Minimodule TQM8315, the latest member 

of the PowerQUICC II Pro module family 

is based on the MPC8315 processor by 

Freescale, and has a few surprises in store. 

Measuring just 70 x 50 mm, TQ offers one of 

the smallest PowerQUICC II Pro modules, 

allowing easy integration into the tiniest of 

spaces. 

News ID 10862 

n Eurotech: embedded computers for 

automated fare collection system 

Eurotech announces a contract to supply embedded 

computers to Cubic Transportation, 

a solution provider of automated fare collection 

systems for public transport. Destined 

for use in the smart card ticketing system on 

the transport system of a major European 

capital, the contract will see 20,000+ units 

installed in buses and station gates throughout 

the network. Each unit will interface directly 

to a smart card reader and relay both passenger 

and journey information into the fare 

collection system. 

News ID 10836 

PRODUCT NEWS 

n Bicker: industry power supplies with 

low standby consumption 

Bicker Elektronik presents the new PC power 

supplies BEA-540H and BEA-550H. The devices 

combine high energy efficiency with the 

robustness of an industry power supply. They 

meet the requirements of the 80 PLUS standard 

and therefore achieve more than 80% efficiency 

at 20 %, 50 % and 100 % load. At 400 and 500 

watts output rating resp., the BEA-540H is 

characterized by 85% and the BEA-550H by 

86% efficiency. 

News ID 10888 

n N.A.T.: AMC-CPU-board with PCIe, SRIO, 

GbE, USB, RS232 and ITDM 

N.A.T. announces the NAMC-8569-CPU, a 

multi-service CPU board featuring multiple 

Ethernet, Serial Rapid IO, PCIexpress, USB 

interfaces and optional TDM connectivity. 

The NAMC-8659-CPU is equipped with two 

processing resources which are the MPC8569 

PowerQUICC III CPU and a powerful FPGA 

from Lattice. 

News ID 11075 

n ADLINK: PCI Express frame grabber 

offering video streaming in full 1080p HD 

ADLINK has released the HDV62, a PCI Express 

frame grabber offering uncompressed image acquisition 

and video streaming in full 1080p HD. 

The HDV62 provides 1920x1080p resolution, 

progressive scan, and noise reduction for greater 

image quality, as well as a wide aspect ratio that 

is more comfortable to the human eye. 

News ID 10808 

n Schroff: Card-Loks supersede torque 

wrench 

Schroff presents the new Card-Lok series 223 

and 224 board retainers from the product 

range of sister company Calmark. Card-Loks 

are used to secure PCBs in place. The new 

series includes a torque limiting feature, for 

which a patent has been filed, so that PCBs 

can be mounted into a chassis with a predetermined 

clamping force and without the need 

for a torque wrench. 

News ID 11000 

n Nexcom: quad head digital signage 

player 

NDiS 542 by Nexcom is a quad head media 

player, which can support four independent 

DVI/Audio outputs. Powered by Intel Core2 

Duo/Core2 Quad series processors and S3 

4500E GPU, NDiS 542 can smoothly playback 

even the most graphic intensive of media including 

full HD video. With four independent 

DVI and audio outputs, NDiS 542 can be utilized 

in a variety of multi-screen deployments 

which would usually require four separate HD 

media players. 

News ID 10931 

43 September 2010

in PDF Format - Embedded-Control-Europe.com

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?