3 ms·
Google used reinforcement learning to design next-gen AI accelerator chips
- adipandas 5y agoI think recently RL or DL is going crazy with design. This is very nice to see. https://www.nature.com/articles/d41586-021-01515-9 https://www.nature.com/articles/d41586-021-01515-9