Are there any detailed examples or case studies demonstrating the use of L2 persistent memory, particularly in GEMM operations?
I actually do not believe that L2 persistent memory would help much with matrix operations as you can alternatively give L2 hints, whether the data should be streamed or is reused. There is probably no added advantage with L2 persistent memory.
1 Like
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.