Analog Compute Platform to Accelerate Edge AI/ML Inferencing
Addressing AI-based computing at the edge and its associated inferencing algorithms, Intelligent Hardware Korea (IHWK) is developing a neuromorphic computing platform for neurotechnology devices and field programmable neuromorphic devices. Microchip Technology, via its Silicon Storage Technology subsidiary, is assisting with the development by providing an evaluation system for its SuperFlash memBrain neuromorphic memory solution.
Based on Microchip’s SuperFlash nonvolatile memory technology and optimized to perform vector matrix multiplication for neural networks through an analog in-memory compute approach, the memBrain evaluation kit enables IHWK to demonstrate the power and efficiency of its neuromorphic computing platform for running inferencing algorithms at the edge.
The memBrain solution stores synaptic weights in the on-chip floating gate in ultra-low-power subthreshold mode, while using the same memory cells to perform the computations, offering significant improvements in both power efficiency and system latency. Compared to traditional digital DSP and SRAM/DRAM based approaches, it delivers 10 to 20 times lower power usage per inference decision and can significantly reduce the overall bill of materials. The final APU is expected to optimize system-level algorithms for inferencing and operate between 20-80 TeraOPS per watt.
Related links:
Microchip Technology
Silicon Storage Technology
MemBrain neuromorphic memory solution