Blending DSP and ML Features into a Low-power General-purpose Processor

With increasing signal processing requirements in various types of IoT and embedded systems, new chips are using both a separate digital signal processor (DSP) and a general-purpose processor to address these increased processing demands. While these suit high-performance devices where silicon area and power are less of a concern, they might not fit small embedded devices that could be difficult for microcontroller software developers to program.

Arm has been working on technologies that boost the signal processing and machine learning capabilities without the pain by combining them into one single processor solution. And recently, Arm announced the new Arm Cortex-M55 processor to take efficient on-device processing to the next level and simplify software development so billions more can deploy endpoint AI.

Read this white paper to learn:

  • How the Arm Cortex-M55 processor, Arm's most AI-capable Cortex-M processor and the first to feature Arm Helium vector processing technology compares to features found in traditional DSPs
  • The fundamental differences between Very Long Instruction Word (VLIW) architecture and the Helium approach to the processor's pipeline design
  • How the processing requirements affect the processor's level-one memory system design and system design considerations

By Joseph Yiu, Distinguished Engineer at Arm.
As presented at Embedded World 2020.