ARM for embedded? Still waiting

By Rich Nass

Executive Vice President

Embedded Computing Design

September 03, 2014

ARM for embedded? Still waiting

How embedded is ARM, really?

Yes, I know there are lots and lots of examples of where ARM-based processors are deployed in embedded applications. But I’d argue that it hasn’t become mainstream and displaced the microcontroller- or Intel-based platforms that sell in the gazillions.

I won’t ask you what an embedded system is, because that’ll take us down a rat hole that’s better left for another time. If you ask ARM, it categorizes embedded as “all applications that use a microcontroller or microprocessor to execute dedicated tasks on its own or within a larger system. Examples include digital signage, household appliances, HVAC systems, engine management, smart metering platforms, touchscreen controllers, and motor control.” For this discussion, that definition will work just fine.

Clearly ARM dominates in the mobile phone segment, and you could certainly argue that that counts as embedded (I’d probably make that argument myself). But for the sake of this discussion, let’s assume that segment doesn’t count. If you’re still with me, is there a place for the ARM architecture in the embedded world?

If you read between the lines of what ARM is offering, the answer would certainly be yes. At the low end, the company made a good pitch to replace the inexpensive microcontroller. While the Cortex-M0 and even some of the higher performing M-based cores have a done a great job, it’s still tough to displace a part that costs just pennies and some surrounding discrete logic.

At the higher end, such as compute-intensive servers, the 64-bit Cortex-A series has some potential for the embedded space. The recent announcement from AMD, which frankly was a surprise to me, adds fuel to that fire. AMD is offering a developer kit, the Opteron A1100, which features ARM’s 64-bit Cortex-A57 core. While the processor isn’t available yet, software developers can start writing their kit with the kit. And the software community is doing its part as well. They are clearly getting ready for the wave of server-based products as evidenced by the push in the Linux camp, including Red Hat, one of the Linux leaders.

As expected, the AMD processor will run the ARMv8 instruction set. When the processor is production ready, models with four and eight cores will become available, as well as all the memory configurations and peripherals you’d expect.

The ARM 64-bit architecture in general lends itself to good performance efficiency relative to the Intel architecture. Hence, a good landing spot is in low-power servers. It’s still too early to see how big that market will get, but the potential is there if all the pieces fall properly into place. With respect to AMD, it’s not surprising that the company is looking at the ARM architecture as a way to differentiate itself from the competition, particularly Intel (this is also discussion that I’ll save for later).

In terms of new designs, industry insiders claim that Intel is winning, but the margin is shrinking as the number of ARM-based embedded designs is growing at a faster clip. Note that there are quite a high number of legacy systems that are based on older architectures, particularly in spaces like high-end imaging, avionics, industrial, and process control. That said, the future for ARM in embedded is bright. It’s just coming along a little slower than some people expected.

Rich Nass, Embedded Computing Brand Director

Richard Nass’ key responsibilities include setting the direction for all aspects of OSM’s ECD portfolio, including digital, print, and live events. Previously, Nass was the Brand Director for Design News. Prior, he led the content team for UBM’s Medical Devices Group, and all custom properties and events. Nass has been in the engineering OEM industry for more than 30 years. In prior stints, he led the Content Team at EE Times,, and TechOnLine. Nass holds a BSEE degree from NJIT.

More from Rich