merge from public

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
+$EXE -t=333 -e=99 -k=13
+$EXE -t=128 -e=32 -k=5 -moe_buf_size=262144

example/ck_tile/15_fused_moe/README.md

@@ -8,6 +8,9 @@ The benifit of this fused-moe:
 * much less kernel instance, easy to maintain
 
 # Implementation and feature support
+## NOTES:
+currently gate+up in fp16 case will very easily cause accumulator overflow the fp16 max(65504), hence result in INF. Please use BF16 for gate+up case, API side will have no check for this.
+
 ## moe-sorting
 this is a common pre-process step before the actual moe-gemm. The purpose is to transform the moe loop over from token-by-token to expert-by-expert, make sure very workgroup is working for a single expert (B matrix). Besides, we extend this op to do the zeroing of the output buffer(to be used for reduce buffer with atomic)