Choosing Between the Intel® Xeon® D Processor and Intel® Xeon® Processor E5 for Embedded Applications
Choosing the right Intel® Xeon® processor can be a daunting endeavor. Intel sells a huge range of processors with varying architectural features, clock speeds, core counts, and thermal dissipation profiles.
But for rugged applications, the choice has become easier with the introduction of the Intel Xeon processor D family. These chips are the first Intel Xeon processors to support an extended temperature range of -40° C to 85° C, making them suitable for extreme environments.
In contrast, standard server processors don’t support extended temperature ranges. Therefore, systems that use these processors may have to incorporate heaters or complex startup routines when operating in cold environments. In hot environments, complex cooling systems may be required.
Alternately, manufacturers might be forced to manually test the temperature range of each processor, to ensure the CPU they ship can handle the full range of operating conditions across its environment.
The Intel Xeon processor D family is also designed for space efficiency. It is a fully integrated SoC, with USB connectivity, SATA, and dual 10 gigabit Ethernet ports integrated on-die. In contrast, standard server processors implement these features in the chipset.
A good point of comparison is the Intel® Xeon® processor E5. Curtiss-Wright performed an exhaustive comparison between the two families that considers size, weight, power, and cost factors (SWaP-C).
At first glance, the Xeon processor E5 Family Rank Margining Tool Calculator may seem like the best choice for demanding applications, as it offers a higher absolute level of performance. A careful analysis often favors the Xeon D processor.
For one thing, Xeon processor E5 has higher TDPs. A Xeon processor E5 uses 75W per-processor, meaning a dual Xeon system in a 6U OpenVPX chassis needs more aggressive cooling than a conduction cooling system can provide.
Xeon processor E5 also uses LGA (Land Grid Array) sockets, which are not designed for rugged environments and may require additional testing before they can be deployed. Xeon D processor, in contrast, uses a soldered BGA (Ball Grid Array) for CPU – motherboard attachment and is considered more robust.
Curtiss-Wright’s white paper doesn’t just discuss these issues in depth; it offers a complex multi-variable equation for evaluating whether a given solution will meet SWaP-C requirements and specific guidelines for how to estimate module and subsystem development costs, as well as equations for estimating system complexity and reliability costs.
Ultimately, Xeon D processor is simply better suited to a wide range of military and rugged applications than its Xeon Processor E5 counterparts. Intel’s embedded platform offers more resilient hardware, fits better into tight TDP envelopes, gives hardware designers more flexibility, operates under a wider range of temperatures, and offers almost the same range of CPU core counts and clock speeds as the latest mainstream Intel Xeon CPU.
