compiler inserts arrive and wait for each GMMA instruction #1645
Unanswered
DeMoriarty
asked this question in
Q&A
Replies: 1 comment 13 replies
-
when you compile this kernel, you must be getting some warnings from ptxas about serialization of the WGMMA instructions. What does it say? |
Beta Was this translation helpful? Give feedback.
13 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
-
In a GEMM kernel that I'm modifying, I noticed that each
HGMMA
instruction is being waited upon immediately:But in the CUDA source code, the HGMMA instructions are committed in batches:
What might be causing the compiler to insert these
DEPBAR.LE
&ARRIVE
?Beta Was this translation helpful? Give feedback.
All reactions