Optimizing multicore SFFs for rugged military environments

Military applications present engineers with some of the most intense design challenges in the embedded world. Not only do modern military designs demand the ability to withstand extended temperature ranges and high/low pressure (altitude and underwater depth) changes, but their computer boards must also withstand severe shock and vibration elements in these harsh environments – all without sacrificing performance. With the growing need for mobility in the military environment, these applications are now increasingly also requiring reduced Size, Weight, and Power (SWaP) to ensure reliable portability. This eliminates the option to increase the footprint to make room for more power or more performance. To compound the challenge, engineers must meet these stringent requirements often with severe time-to-deployment and budgetary constraints.

At the processor level, the advent of multicore technology has proven to be an attractive option to help conquer these challenges, since it offers the ability to scale applications and add features within embedded form factors without dramatically affecting the energy variables such as thermal output and power consumption.

Explained most simply, multicore processor architecture places two or more processor-based "execution cores" within a single processor. This multicore processor plugs directly into a single processor socket, but the operating system perceives each of its execution cores as a discrete logical processor, with all the associated execution resources. The idea behind this implementation of the chip's internal architecture is in essence a "divide and conquer" strategy. By divvying up the computational work performed by the single microprocessor core in traditional microprocessors and spreading it over multiple execution cores, a multicore processor can perform more work within a given clock cycle. For instance, engineers can integrate controls that previously required separate dedicated systems into one system. This means that only one computer is needed for control and visualization tasks – even for critical and highly complex real-time applications.

Accordingly, designers must be careful to optimize multicore by selecting key system-level features. A strong knowledge of the several SFFs available on the market today also enables designers to utilize multicore effectively.

Technologies to optimize multicore processors

With the proliferation of multicore offerings, it is important to keep in mind that not all multicore platforms are created equal. To maximize the capabilities of multicore technologies, a number of key system-level features must also be available. The platform approach from Intel for multicore processing combines a multicore architecture with complementary system-enhancing technologies to enable developers to leverage the architecture very efficiently. These multicore technologies include:

• Thread-level parallelism ‚Äì In order to take full advantage of multicore processing performance, both the operating system and applications running on the computing platform must support a technology called thread-level parallelism. A processor equipped with thread-level parallelism can execute completely separate threads of code. This allows one thread running from an application and a second thread running from an operating system, or parallel threads running from within a single application.
• Hyper-Threading Technology ‚Äì Multicore capability reflects a shift to parallel processing ‚Äì a concept originally conceived in the supercomputing world. Hyper-Threading Technology (HT Technology) enables processors to execute tasks in parallel by weaving together multiple "threads" in a single-core processor. Whereas HT Technology is limited to a single core's using existing execution resources more efficiently to better enable threading, multicore capability provides two or more complete sets of execution resources to increase compute throughput. This increases the amount of work a processor can do in the same time as a processor with one core. With HT technology, one dual-core processor is able to simultaneously run four software threads. As more processors are added to a server, the number of supported threads increases to help deliver better overall performance.
• Virtualization ‚Äì Another important feature is Virtualization Technology (VT), which allows multiple operating systems and applications to run as "virtual machines" in independent partitions on one platform with simpler hardware administration. This makes overall systems more stable because processes that would collide on single-core systems can be separated. The partitions can be assigned as necessary, even when the system is running. VT offers the option to integrate previously stand-alone systems such as controllers, firewalls, and data servers, completely isolated from each other in a single system. Multicore and VT together offer innumerable configuration possibilities and a degree of freedom for implementing multiple applications on one system, which ultimately leads to savings in hardware.

Embedded SFFs utilizing multicore

Until recently, VME was the solution of choice for high-end military applications. However, current VME designs are unable to meet some of the new application demands due to beyond-budget price points, high power dissipation, and relatively large size. To satisfy the need for smaller size and lower cost, a number of standards-based, modular, off-the-shelf rugged form factors have emerged and are utilizing multicore processing technology: COM Express, ETX, CompactPCI, AdvancedTCA, and MicroTCA. Each has its own variables of size, performance, power dissipation, and price points to consider, providing a wider range of options for military applications. (See Table 1 for a comparison.)

CompactPCI has thrived in the military market, since the limitations of VME-based architectures have been unable to keep up with the requirements. The 6U form factor has quickly replaced VME in large custom designs. But as the pressure to reduce size and weight intensifies – particularly for Unmanned Aerial Vehicles (UAVs) that carry an increasing array of electronics – the smaller 3U CompactPCI is gaining popularity. The 3U form factor also offers ruggedization benefits due to its significantly greater stiffness, making it less susceptible to shock and vibration.

AdvancedTCA, with its high bandwidth and design flexibility, has also made significant in-roads within military communication applications that are not in the conflict zone. Targeted to the requirements of next-generation carrier-grade communications equipment, AdvancedTCA incorporates the latest trends in high-speed interconnect technologies, next-generation processors, and improved reliability, manageability, and serviceability.

MicroTCA is a relatively new PICMG standard for open modular systems and preserves many of the important philosophies of AdvancedTCA, including the basic interconnect topologies and management structure. Whereas AdvancedTCA is optimized for very high-capacity applications, MicroTCA is designed to address cost-sensitive and physically smaller applications with lower capacity and performance requirements. By configuring highly diverse collections of AdvancedMCs in a MicroTCA shelf, many different applications can be easily realized.

The Kontron OM6040 is a compact and modular MicroTCA platform that is ideally configured for the design of small, compact, and highly integrated multiprocessor systems based on either PCI Express (PCIe) or GbE packet switching backplane technologies. The small, packet-switched backplane system is an optimal choice for telecommunication applications for demanding military networking environments including interoperable integrated MicroTCA and WiMAX solutions for fixed wireless, last-mile wireless, back-haul capabilities, and cellular for military vehicles.

Embedded Technology eXtended (ETX) has established itself as a popular non-backplane form factor. Targeting customizable embedded requirements, ETX modules offer reliable operation and a long life even in harsh environments. ETX modules employ heat-spreader plates to create a larger surface area over the module to assist with conduction cooling, making these COM modules an option for use in extended temperature ranges as long as the components in the design can tolerate the harsh environment.

ETX modules have been successfully implemented into rugged ultra-mobile PC applications that demanded a modular design with great flexibility. The photo on page 28 shows a small, rugged, and portable device that was designed quickly using a semi off-the-shelf solution along with custom BIOS. In this application, ETX modules, like the one shown in Figure 1, provided maximum performance that allowed the mobilized system to fulfill the demand for mission-critical, high-end computing.

Figure 1

ETX modules have also proven successful in avionics where shock and vibration are among the greatest issues. Through the small, space-saving design of an ETX CPU module, many space-restricted applications can benefit. ETX modules, specifically the ETX-CD and ETX-PM derivatives, have also proven ideal for custom designs involving computing modules in mobile platforms.

Small and rugged, Computer-on-Module implementations are ideal for a broad range of embedded applications where they fit mechanically, economically, and functionally, and where other form factors such as add-in cards cannot be used. High-performance segments can use COM Express modules to help transition designs reliant on legacy bus technologies to future-focused technologies such as PCI Express and Serial ATA. The form factor flexibility of COM Express with its five pin-out types enables developers to segment their designs for different classes of embedded applications. Applications such as unmanned vehicles, training simulators, and portable tactical communications devices can all benefit from the COM Express form factor.

Multicore SFFs meet current and future needs

Multicore processor-based small form factors offer a variety of attractive choices for today's rugged military applications with their ability to meet the military's increasing SWaP demands, while also holding a promising future for evolving requirements. Accordingly, thread-level parallelism, hyper-threading technology, and virtualization are key to this equation. As a result, the number of architectures that can stand up to the demands of military applications has grown and offers many performance, size, and cost alternatives to choose from. As new technologies come to market, even more powerful computing platforms are on the horizon.

Christine Van De Graaf is the product marketing manager for Kontron America's Embedded Modules Division located in Northern California's Silicon Valley. She has more than seven years of experience working in the embedded computing technology industry and holds an MBA in Marketing Management from California State University, East Bay, Hayward, CA. Christine can be contacted at Christine.vandegraaf@us.kontron.com or 510-661-2220, ext. 250.

David Pursley is a field applications engineer with Kontron. He is responsible for business development of Kontron's MicroTCA, AdvancedTCA, CompactPCI, and ThinkIO product lines in North America and is based in Pittsburgh, PA. Previously, he held various positions as a field applications engineer, technical marketing engineer, and marketing manager. David holds a Bachelor of Science in Computer Science and Engineering from Bucknell University and a Master's degree in Electrical and Computer Engineering from Carnegie Mellon University. He can be contacted at david.pursley@us.kontron.com or 412-919-3709.