apache / tvm

Open deep learning compiler stack for cpu, gpu and specialized accelerators

Home Page:https://tvm.apache.org/

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

[Tracking Issue] Scalable Matrix Extension (SME) upstreaming

lhutton1 opened this issue · comments

This issue is to track progress for SME enablement upstream

  • P0. Fixed Virtual Platform (FVP) and toolchain installation - #16755
  • P1. Fixed Virtual Platform (FVP) functional testing infrastructure - #16749
  • P2. Target parser support for SME - #16794
  • P3. Processor state annotations codegen support - #16761
  • P4. Scalable FP32 Dense SME schedule - #16921
  • P5. Scalable FP32 Conv2d SME schedule - #17003
  • P6. Scalable FP16->FP32 Dense SME schedule - #16981
  • P7. Scalable FP16->FP32 Conv2d SME schedule
  • P8. Optimize tiling of tensor intrinsics based on cache size