Merge branch 'gfx950' into lwpck-2619

efab74a3 · Rostyslav Geyyer · 86950b3a · bcef33c1 · efab74a3 · efab74a3
Commit efab74a3 authored Jan 24, 2025 by Rostyslav Geyyer
20 changed files
--- a/example/ck_tile/12_smoothquant/example_smoothquant.cpp
+++ b/example/ck_tile/12_smoothquant/example_smoothquant.cpp
--- a/example/ck_tile/12_smoothquant/instances/smoothquant_instance_common.hpp
+++ b/example/ck_tile/12_smoothquant/instances/smoothquant_instance_common.hpp
--- a/example/ck_tile/12_smoothquant/smoothquant.cpp
+++ b/example/ck_tile/12_smoothquant/smoothquant.cpp
--- a/example/ck_tile/12_smoothquant/smoothquant.hpp
+++ b/example/ck_tile/12_smoothquant/smoothquant.hpp
--- a/example/ck_tile/13_moe_sorting/moe_sorting_api.cpp
+++ b/example/ck_tile/13_moe_sorting/moe_sorting_api.cpp
--- a/example/ck_tile/13_moe_sorting/script/smoke_test.sh
+++ b/example/ck_tile/13_moe_sorting/script/smoke_test.sh
@@ -16,4 +16,5 @@ $EXE -t=127 -e=99 -k=19
 $EXE -t=71 -e=11 -k=11
 $EXE -t=1 -e=1 -k=1
 $EXE -t=99 -e=2 -k=1
 $EXE -t=333 -e=99 -k=13
\ No newline at end of file
+$EXE -t=128 -e=32 -k=5 -moe_buf_size=262144
--- a/example/ck_tile/14_moe_smoothquant/instances/moe_smoothquant_instance_common.hpp
+++ b/example/ck_tile/14_moe_smoothquant/instances/moe_smoothquant_instance_common.hpp
--- a/example/ck_tile/14_moe_smoothquant/moe_smoothquant.cpp
+++ b/example/ck_tile/14_moe_smoothquant/moe_smoothquant.cpp
--- a/example/ck_tile/14_moe_smoothquant/moe_smoothquant.hpp
+++ b/example/ck_tile/14_moe_smoothquant/moe_smoothquant.hpp
--- a/example/ck_tile/15_fused_moe/README.md
+++ b/example/ck_tile/15_fused_moe/README.md
--- a/example/ck_tile/15_fused_moe/fused_moe.hpp
+++ b/example/ck_tile/15_fused_moe/fused_moe.hpp
--- a/example/ck_tile/15_fused_moe/fused_moegemm.hpp
+++ b/example/ck_tile/15_fused_moe/fused_moegemm.hpp
--- a/example/ck_tile/15_fused_moe/instances/fused_moe_api.cpp
+++ b/example/ck_tile/15_fused_moe/instances/fused_moe_api.cpp
--- a/example/ck_tile/15_fused_moe/instances/fused_moegemm_api.cpp
+++ b/example/ck_tile/15_fused_moe/instances/fused_moegemm_api.cpp
--- a/example/ck_tile/15_fused_moe/instances/fused_moegemm_api_internal.hpp
+++ b/example/ck_tile/15_fused_moe/instances/fused_moegemm_api_internal.hpp
--- a/example/ck_tile/15_fused_moe/instances/fused_moegemm_api_traits.hpp
+++ b/example/ck_tile/15_fused_moe/instances/fused_moegemm_api_traits.hpp
--- a/example/ck_tile/15_fused_moe/instances/fused_moegemm_bf16_m32.cpp
+++ b/example/ck_tile/15_fused_moe/instances/fused_moegemm_bf16_m32.cpp
--- a/example/ck_tile/15_fused_moe/instances/fused_moegemm_fp16_m32.cpp
+++ b/example/ck_tile/15_fused_moe/instances/fused_moegemm_fp16_m32.cpp
--- a/example/ck_tile/15_fused_moe/instances/fused_moesorting_api.cpp
+++ b/example/ck_tile/15_fused_moe/instances/fused_moesorting_api.cpp
--- a/example/ck_tile/15_fused_moe/main.cpp
+++ b/example/ck_tile/15_fused_moe/main.cpp