Reorder global load instructions of dtva custom kernel

ROCm / hipBLASLt

hipBLASLt is a library that provides general matrix-matrix operations with a flexible API and extends functionalities beyond a traditional BLAS library

https://rocm.docs.amd.com/projects/hipBLASLt/en/latest/index.html

MIT License

49 stars 80 forks source link

Reorder global load instructions of dtva custom kernel #869

Closed briannwu closed 3 months ago

briannwu commented 3 months ago

Custom_Cijk_Alik_Bljk_BBS_BH_Bias_AS_SAV_UserArgs_MT256x256x64_MI16x16x1_SN_K1_MIWT4_16_DTVA

hcman2 commented 3 months ago

Will you add a test yaml to run this kernel?

briannwu commented 3 months ago

Will you add a test yaml to run this kernel? added