Nvidia has been working on a prototype multi-die AI accelerator chip called RC 18. The 36-module strong chip, developed by Nvidia Research, is currently being evaluated in the labs, and its highly scalable, interconnected design could act as precursor to high-end, multi-GPU graphics cards.
The current prototype for Nvidia’s multi-die solution was taped out on TSMC’s 16nm process – the same one utilised in most 10-series GeForce graphics cards. Its rather small footprint is made up of 36 tiny modules, each one comprised of 16 PEs (Processing Elements), a basic RISC-V Rocket CPU core, buffer memory, and eight GRS (Ground-Referenced Signaling) links totally 100GB/s I/O bandwidth per chip. All in all the chip is fitted with some 87 million transistors in total.
To read more, please visit: https://www.pcgamesn.com/nvidia/nvidia-rc-18-ai-deep-learning-accelerator-mcm-multi-gpu?amp.