Embedded Executive LIVE
summary
Brainchip describes an AI edge co‑processor claiming 200 mW full power operation with cores that scale to sub‑milliwatt cores and a pico core in the 0.1 mW range that can run a coin‑cell for a year. Toolchain converts TensorFlow/ONNX CNNs to spiking neural network conversion and performs fixed‑point quantization. Production chips are shipping in vibration and audio detectors, seizure‑prediction smart glasses, and PCIe/Raspberry Pi dev boards — customers report deployed detections waking higher‑power hosts.
AI-written summary · about
related talks
20mLow Power AI Edge Hardware
55mEdgeAI - AI in the Palm of Your Hand
21mUnlocking the potential of low-power embedded devices with Edge AI software
22mAI On The Farthest Edge
13mHow Edge AI Will Save The World
28mML on the Edge for Industry 4.0 with Arm Ethos N-78 Neural Processing Unit
37mDeep Learning Neural Network Acceleration at the Edge
11mRISC-V enabled, low-power CNN classification in Edge devices