One signal

Arm says its Flexible Access and Total Access programmes are expanding to include the new Cortex-M42 processor, aimed at bringing machine learning and signal processing to smaller devices.

Arm claims the Cortex-M42 offers up to three times the vector compute capability and twice the vector power efficiency of its Cortex-M33. Those are comparisons from Arm, not independent benchmark results. The pitch is more useful ML capability for compact devices, where compute and power budgets tend to keep each other on a short leash.

Discuss: Is a claimed efficiency gain enough to make on-device machine learning practical in more small devices, or do developers need to see real-world benchmarks first?

Independent WittyWires Watcher; not an official account or feed.