Skip navigation

Embedded system designers have many fewer processor and chipset options than colleagues designing enterprise class systems. AdvancedTCA is a great platform for NEBS compliant, carrier grade systems but the 200W per blade power envelope further restricts the options for system designers. So the introduction of new processors and chipsets with lower power and extended life cycles, that have been developed with ATCA in mind, is an important development that opens up new opportunities for system designers.


With the introduction of the Quad-Core Intel® Xeon® Processor 5400 Series embedded performance has moved to the next level. These 45nm devices not only integrate twice the number of cores found in a dual core 5100 series processor but also 50\% more cache per core, further increasing performance. As discussed in a recent Webinar hosted by Light Reading quad core processors can dramatically increase system performance, especially in embedded applications which are usually well suited to parallel execution.


For embedded systems in general, and ATCA based systems in particular, performance per watt is a key parameter. This metric needs to be applied to a complete board implementation not just the processor, so chipset and memory are also important. The Intel® 5100 MCH chipset delivers significant power savings with DDR2 memories instead of power hungry FBDIMMS. Combining the latest quad core processors and chipsets delivers increases in performance per watt of over 70\%.



According to a recent report on ATCA from Heavy Reading there are already five vendors with dual-core Intel® Xeon® Processor based ATCA blades. With the significant increase in performance per watt offered by the combination of Quad-Core Intel® Xeon® Processor 5400 Series, Intel® 5100 MCH chipset and DDR2 memory we can expect to see these vendors, and more, bringing out ATCA blades with two quad core processor devices, delivering greater than 2x performance within the 200W power envelope.



Message Edited by serenajoy on 03-11-2009 08:16 PM

Looking at the main trends for CPU and chipset development, we see performance increasing but board and chip sizes getting smaller and smaller, this change is driven by popular demand for smaller embedded form factors. Smaller form factors help system designers integrate their solutions into a wider variety of applications. Take the advertising industry; display advertising is powered by ever-larger digital signage screens that require a robust back-end infrastructure to deliver rich media content in public spaces. Retail kiosks need powerful CPU performance to distribute engaging content and interactive solutions at the point of information and sale.


Intel 945GME-based platforms support a range of powerful multi core engines and rich I/O features. They are perfectly suited for Advantech's reliable 3.5" Biscuit embedded SBC platforms. Dual Independent displays and unique Module I/O (MIO) interface answers the need for powerful and flexible 3.5" SBCs requiring high resistance to vibration, and plenty of expansion possibilities. MIO integrates the most popular bus signals together into a high-density 160-pin connector. These bus signals include PCI, High-Speed USB 2.0, Digital Video Output (DVO), Low Pin Count (LPC), and AC97 audio.

King Demandius tasked his two royal advisors, Max and Rupert, with fixing the problems with the road between the kingdom's two most prosperous cities: Hither and Yon. Max, the head of the ministry of taxes and tolls, realized that royal revenue could be improved if traffic on the road was increased. Horse and chariot speeds had already been pushed to their limits. His solution: widen the road in order to allow multiple travelers to pass simultaneously. Max's proposal was warmly accepted by the king, giving rise to the kingdom's first multilane-tollway.


Rupert, on the other hand, had received complaints from the king that travel time between the royal palace in Hither and the vacation home in Yon was just too long. During these trips the road was closed to everyone but the royal caravan so additional lanes were of no benefit. Horse and chariot speeds couldn't be increased. What was Rupert to do? In the end he made two proposals: Move Hither closer to Yon or add in-chariot entertainment to the royal caravan to help pass the time. Rupert now enjoys full accommodations in the dungeon.


In the same way, it is the best of times and the worst of times for embedded processor users. If you are looking to increase the throughput of your application, like Max, adding more cores is equivalent to creating a multilane expressway. More cores can be a very effective solution for increasing total application capacity. However, if your problem is more like Rupert's, your options are a bit more limited. Core speeds are not increasing and more core's by themselves, like more highway lanes, might do very little to improve application speed.


Fortunately, there can be a happy ending to our story. Although less obvious, most multicore processors do offer possibilities for improving application speed. I'll be talking about some of these on November 13 in a webinar hosted by Open Systems Publishing. I'd love to have you there. In the meantime, if you have any success stories or tricks to increase application speed on Intel multicore processors, I invite you to share them here.

Sure, this little booklet is really intended for 'newbies' who want to learn more about what COM Express, its history and how it's used. But it's a good 'read' and emphasizes several areas that are so vital to working with COM Express. For example, COM Express helps you maintain a broad product portfolio by enabling you to follow a common-platform strategy in your design. MODULARITY: you can offer low, medium, and high-performance versions simply by using different levels of COM Express modules on an otherwise common carrier. Two areas where COM Express has already helped to build broad product portfolios include: medical and instrumentation.


Medical systems are highly sophisticated and complex; using a single shared custom carrier design, COM modules of different performance levels makes the upgrade path easy. This makes everyone happy and is cost-effective. At RadiSys, we love working with our customers in a forward-thinking way - we work as a 'design team' to make a COM design successful, providing the software utilities, BIOS features, critical schematics, and mechanical diagrams so that they accelerate their time to market.


So this little booklet, as basic as it may seem, can provoke BIG thoughts and BIG ideas...take it with you on your next plane ride or read it while eating lunch...enjoy!



Get your free copy!


Lorraine Orcino

Sr. Product Line Manager

Message Edited by serenajoy on 03-11-2009 08:04 PM

Which came first?

Posted by kontron_cvdg Aug 19, 2008

I doubt I'm the only one who's pondered the question of "which came first: the chicken or the egg". And I might not even be the first person to consider how that might compare to small form factor computer-on-modules and SBCs and ever-shrinking multi-core processors. But, go with me on this one.


ETX (Embedded Technology eXtended) got its start in the year 2000 and initiated the move from full custom motherboards to standard, off-the-shelf computer-on-modules with customizable carrier boards. Then, along came COM Express in around 2003 and PC/104-sized, "micro" COM Express compatible modules as well. If my memory is correct, the Intel® Tick-Tock strategy really took off around the year 2005, leading to the release of the first multi-core processors not long there after.


Although the smallest Computer-On-Modules haven't been released with dual and quad-core processors yet, it's a safe bet to say that they are definitely on the way. And, it's foreseeable that their demands will push for even further tick-tock silicon technology advancements.


What do you think - Small form factor module or small form factor multi-core architecture: Which came first?






Christine Van De Graaf



Product Marketing Manager, Embedded Modules Division





Message Edited by serenajoy on 03-11-2009 08:39 PM

…when you’re just back from vacation, reading this blog and wondering how you’re going to architect your next project. You’ll have heard of AdvancedTCA and MicroTCA but are they suitable? Maybe your application is not telecom related but does have a communication aspect?


Well, for a relatively minor outlay there are a number of events coming up that will enable you to see the rapidly expanding range of Intel x86 based processor AMCs and xTCA systems available. In addition to tabletop exhibitions where you actually get to see and touch the products, you can talk to the technical experts. These events tend to have comprehensive speaking programs covering a variety of compelling topics from keynotes and industry perspectives to detailed technical workshops.


MicroTCA Conference 2008, September 24th at Reading, UK


ATCA/uTCA Special Focus Workshop as part of the 2008 Nuclear Science Symposium, October 18-19th at Dresden, Germany,


AdvancedTCA Summit featuring MicroTCA, October 21-23rd at Santa Clara, CA, USA, Rose Schooler, General Manager of Intel’s Performance Products Division in the Embedded Computing Group will be giving a keynote on Network 2015 along with other industry luminaries such as our own Shlomo Pri-Tal.


Embedded Conference Scandinavia 2008, October 21-22nd at Stockholm, Sweden,


If you want to learn more about xTCA and AMC technology, then attending one of these events can be a very effective use of your time. While I’d like to think you’d come straight to the Emerson Network Power booth, the best thing about all these shows is the diversity of content you will see. There are an amazing number of modules, carriers, systems and other assorted building blocks available today, and it is possible to construct many more solutions off the shelf for fast prototyping and development.


There is one other event worth mentioning, although it is not open to the public, and that is the latest Advanced Interoperability Workshop (AIW) which is being held in Germany during September. The AIW gives engineers the chance to do detailed interop between all the elements of a system including the chassis, power modules, processors and switches and enables them to iron out any issues before making these products available to the public. At some point in the future, the Communications Platform Trade Association will formally certify products. Meanwhile, the work of the AIW continues to make AMC ‘plug and play’ a reality in most instances.


Have a great summer and perhaps I’ll see you at one of these events!

Message Edited by serenajoy on 03-11-2009 08:17 PM

Think Inside The Box

Posted by C_Ruparel Aug 18, 2008

How Nehalem Changes the Game for ATCA Design


Over the last decade, the design engineering community has been conditioned to think that all good ideas come from thinking "outside the box" - that is, to allow creativity to flow, without much regard for practical constraints. In the ATCA design domain, the challenge is how to think creatively inside the box - specifically, how to squeeze more performance into equipment built to the ATCA specification.


As the ATCA market has evolved, handling increasing thermal issues and power requirements has become an exercise in true design creativity. Designing to the ATCA spec, with its fixed power, thermal and form factor, has made it difficult to meet market demand for the combination of:


  • higher CPU performance (which until now was possible only with a dual socket configuration)

  • more memory

  • value-added features such as local storage, AMC, etc.


How Nehalem changes the game


Intel processors' performance keeps pace with Moore's Law, doubling every 18 months or so. But the Nehalem processor does more than deliver higher compute power; it has a new micro architecture with several improvements, particularly in the memory subsystem. Nehalem has an integrated memory controller, twice the memory bandwidth (generation to generation), and more memory channels.


Add hyperthreading (allowing two threads per core) to the mix and Nehalem gives you have eight virtual cores/sockets. This offers better platform performance for a wide array of applications than the eight physical cores from previous processor generations. Go figure -- Nehalem delivers better platform performance than two previous-gen quad-core processors, for the cost of one!


With Nehalem launch around the corner (Q4'08-Q1'09), picking Nehalem for the 2008 ATCA SBC is as simple as picking Olympian Michael Phelps for the 200 meter freestyle swim competition. You get the point.


In short, Nehalem makes it possible to design a high-performance ATCA CPU board possible with just one processor. This is the what’s game-changing for ATCA system designers.


Nehalem enables fundamental design improvements


With its superior performance and feature set, Nehalem will allow you to squeeze more out of the inherent constraints of the ATCA specification. You can leverage Nehalem to meet the market's 2x-3x generation-to-generation performance expectation with just a single CPU on ATCA. This frees up real estate and power on the board to allow for more memory and platform features such as storage and AMC.


Specifically, you can:


  1. Use a single board to address a broader set of memory-intensive applications, because you will have much more memory capability on-board, along with Nehalem’s performance boost.

  2. Deliver a higher-performance system at a lower cost with a single CPU and more DIMM slots. More DIMM slots allow lower-density DIMMs that are cheaper and run cooler than higher density ones.


Reducing risk and total cost of ownership


By packing more capabilities onto a single board, a broader set of applications can be addressed with the same board. This means fewer SKUs to manage and inventory, lower manufacturing costs, and a drop in validation requirements.


RadiSys is committed to providing performance at the beat rate of the silicon. Nehalem presents an exciting array of possibilities for all of us in the embedded community ... inside every ATCA system.


Chandresh Ruparel

Sr. Product Line Manager


Message Edited by serenajoy on 03-11-2009 08:17 PM

Multi-core, multi-display, fully-featured, energy-saving, and mobility? What else are you looking for in a tiny board? Is it be feasible to have such board embedded with all these features and still squeeze them into small form factor? Yes, it is. There is no doubt that applications are getting more and more complex than before, so a small board is not good enough to meet all requirements and its features are limited. The balance between size and functionality is always a dilemma for a product planner. There is no definitive answer between the struggle of what feature must go and what must stayt. Fortunately with Intel's newly launched Montevina platform this problem is resolved.


Advantech's GM45 Express based Mini ITX motherboard is the enable to boost up the embedded applications without any additional modules. The Intel GM45 Express chipset is equipped with rich I/O connectivity and supports a 45nm based Core 2 Duo mobile processor that is well-suited for today's and tomorrow's embedded applications. In addition to analog VGA and LVDS panel support, GM45 provides a cost-effective way to support digital display devices, such as DVI, HDMI, and DisplayPort. This great feature supports a variety of display devices in one PCB design. The highly integrated Mini ITX motherboard has a compact size of 6.69" x 6.69" , and is a truly all-in-one platform that not only resolves space constraint concerns but also fulfills the strong demand for multi-core, rich I/O connectivity and expandability. The Mini ITX motherboard brings more opportunities to employ a small form factor board for performance and feature demanding applications.

It's The Software, $\%*&!

Posted by EricRSYS Aug 15, 2008

The Increasingly Important Role of Commercial Middleware in ATCA


Today, whenever the talk is about ATCA architecture, middleware is increasingly part of the dialog. This is exciting because it means that equipment manufacturers are looking for, and finding, additional ways to deliver value beyond COTS hardware. For engineers, the incorporation of middleware into ATCA devices presents a way to not only accelerate next-generation platform deployment; it also frees up internal resources to focus on their value-added applications.


What to do – Buy commercial? Or go open source?


In the past, the ATCA community has had discussions around middleware standardization, primarily focused in the Service Availability Forum (SAF). However, over the past year, additional attention has been paid to SAF by actual users, rather than just contributors. As with any standard, SAF becomes a key discussion point for equipment manufacturers purchasing middleware.


From both a business and engineering perspective, the most efficient way to address the middleware segment - in a reliable manner that mitigates risk - is to acquire the software from an organization that has a funded, dedicated organization building to the SAF specs. In short, commercial software. There are three reasons for this:


  • First, a for-profit commercial software organization has a vested interest in keeping current with the specifications - that's the only way they will continue to attract new customers. What customer will be satisfied with software that is compliant to a standard more than two years old?


  • Second, because commercial software organizations have a vested interest in seeing customers succeed, they will provide customers with access to technical resources when questions arise. In high-profile ATCA communications equipment, the impact of industrial-grade tech support is crystal clear: If a major carrier experiences a service outage and it's traced to the middleware on your equipment, who's going to fix the problem?


  • This leads to the third reason to use commercially developed middleware: indemnification, a topic that is rarely discussed within the development community but which is extremely important to the suits in the Finance, Legal and Procurement departments. In an increasingly litigious business environment, customers at all levels are asking to be indemnified from any legal issues that may arise from software provided to them.


All together, these reasons make a compelling case for commercial middleware; they show that factors other than cost must be taken under consideration when making a middleware choice.


Eric Gregory

Sr. Product Line Manager


Message Edited by serenajoy on 03-11-2009 08:18 PM

I had the privilege of supporting a recent RadiSys design win for technology used by The People's Republic of China to broadcast the 2008 Beijing Olympics via Internet TV in the southern province of Guangdong. We're proud to be a part of the Olympics - and what a ringing endorsement for the ATCA form factor!


Besides the fact that the Olympics is one of my favorite sporting events, the ability of a customer to take a COTS ATCA platform 2 months prior to deployment and integrated their leading edge applications and integrate them into a live network prior to the event is equivalent to setting a world record 100 meter dash!


The RadiSys ATCA application-ready platform is anchored on Intel Architecture server blades and is ready for expansion as network demands pick up. The Olympics will be delivered like never before - whether you're watching Ping Pong on your mobile phone or real-time baseball streaming from multiple monitors at an Internet Cafe. I imagine this as a decathlon event for ATCA and I am excited by the chance to be a part of bringing the Olympic experience to many of the citizens around the world. If anybody else has an Olympic story built on ATCA, send in to the blog


Keate Despain

Sr. Director

ATCA Product Line Management


Message Edited by serenajoy on 03-11-2009 08:19 PM

Filter Blog

By date: By tag: