VanishD Shuaiwei Huang Allen Wu Weihua Du commited on
Commit
87bcc9f
·
verified ·
1 Parent(s): 48928f0

README: describe the dataset as in the paper

Browse files

Co-authored-by: Shuaiwei Huang <112870732+ArtificialRay@users.noreply.github.com>
Co-authored-by: Allen Wu <110312634+wuallenwu@users.noreply.github.com>
Co-authored-by: Weihua Du <73473861+StigLidu@users.noreply.github.com>

Files changed (1) hide show
  1. README.md +3 -3
README.md CHANGED
@@ -28,9 +28,9 @@ configs:
28
 
29
  # KernelShift
30
 
31
- Benchmark data for optimizing **Arm (AArch64) CPU kernels** across three
32
- generations of vector and matrix extensions: SVE (AWS Graviton3, 256-bit),
33
- SVE2 (AWS Graviton4, 128-bit) and SME2 (Apple M4). Each kernel *definition*
34
  comes with *workloads*, a scalar *reference* that serves as the correctness
35
  check and the agent's starting point, and an *expert baseline*: a candidate
36
  kernel is checked against the reference and timed against the baseline.
 
28
 
29
  # KernelShift
30
 
31
+ Benchmark data for KernelShift: kernel optimization tasks for **Arm (AArch64)
32
+ CPUs** across four generations of vector and matrix extensions: Neon,
33
+ SVE (AWS Graviton3, 256-bit), SVE2 (AWS Graviton4, 128-bit) and SME2 (Apple M4). Each kernel *definition*
34
  comes with *workloads*, a scalar *reference* that serves as the correctness
35
  check and the agent's starting point, and an *expert baseline*: a candidate
36
  kernel is checked against the reference and timed against the baseline.