-
Notifications
You must be signed in to change notification settings - Fork 1.2k
Issues: NVIDIA/cutlass
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Author
Label
Projects
Milestones
Assignee
Sort
Issues list
[QST] uncoalesced shared accesses via Copy_Atom<SM80_CP_ASYNC_CACHEALWAYS<uint128_t>, ElementA>
? - Needs Triage
question
Question
#2202
opened Mar 27, 2025 by
tlogn
[QST] Ampere GEMM 2.x vs 3.x Performance Gap
? - Needs Triage
question
Question
#2201
opened Mar 26, 2025 by
davidpissarra
[QST] cutlass/epilogue/fusion/sm90_visitor_tma_warpspecialized.hpp occur error.
? - Needs Triage
question
Question
#2198
opened Mar 26, 2025 by
matrix97317
[QST] Any example of MainloopSm90ArrayTmaGmmaWarpSpecializedMixedInput
? - Needs Triage
question
Question
#2193
opened Mar 24, 2025 by
ankutalev
[QST] Fusion paths on Ada/SM89
? - Needs Triage
question
Question
#2189
opened Mar 23, 2025 by
jstoecker
[QST] Where MMA operation is performed?
? - Needs Triage
question
Question
#2187
opened Mar 21, 2025 by
IzanCatalan
[QST] Support for GEMM with arch 120
? - Needs Triage
question
Question
#2186
opened Mar 20, 2025 by
jungi-lee
[BUG] Potentially too many registers for softmax in Blackwell FMHA?
? - Needs Triage
bug
Something isn't working
#2183
opened Mar 18, 2025 by
Sinestro38
[QST] Why we have both Ping-pong and Cooperative schedule?
? - Needs Triage
question
Question
#2181
opened Mar 17, 2025 by
Algy
[BUG] Incorrect results with Hopper Mixed-input Kernel using 8x1x1 cluster
? - Needs Triage
bug
Something isn't working
#2176
opened Mar 15, 2025 by
manishucsd
[BUG] Segmentation fault with high N, C, K values in Conv2dFprop
? - Needs Triage
bug
Something isn't working
#2175
opened Mar 14, 2025 by
IzanCatalan
[QST]How can I implement W4A16 gemm kernel with cute api?
? - Needs Triage
question
Question
#2168
opened Mar 13, 2025 by
liuqi123123
[QST] Adding new parameter to Conv2dFprop in Python
? - Needs Triage
question
Question
#2166
opened Mar 12, 2025 by
IzanCatalan
[QST] Variable size Gemm that can be Cuda graphed
? - Needs Triage
question
Question
#2164
opened Mar 12, 2025 by
otropp
[QST] if(runtime_value) (cute::copy or cute::gemm) Generates NaNs
? - Needs Triage
question
Question
#2163
opened Mar 12, 2025 by
phantaurus
[BUG] Possible race condition in sm90_gemm_array_tma_warpspecialized_cooperative
? - Needs Triage
bug
Something isn't working
#2162
opened Mar 11, 2025 by
thefacetakt
[DOC] CUTLASS INT4 GEMM: Missing SM89 Dispatch Configuration for L40S/4090
? - Needs Triage
documentation
Documentation
#2158
opened Mar 8, 2025 by
hxu296
[BUG] BF16 * BF16 GEMM with pingpong schedule and non-zero beta hangs
? - Needs Triage
bug
Something isn't working
#2152
opened Mar 6, 2025 by
manishucsd
[BUG] CUTLASS Python Interface nvrtc fails on Hopper
? - Needs Triage
bug
Something isn't working
#2150
opened Mar 5, 2025 by
sommerlukas
[BUG] Possible numerical problems with a warpspecialized_cooperative_epi_tma op
? - Needs Triage
bug
Something isn't working
#2147
opened Mar 4, 2025 by
henrylhtsang
[QST] which is optimised way to iterate over the conv2d filter tensor
? - Needs Triage
question
Question
#2146
opened Mar 4, 2025 by
IzanCatalan
[QST] Bitwise Operations with Cutlass datatypes
? - Needs Triage
question
Question
#2145
opened Mar 4, 2025 by
IzanCatalan
[QST] Why does GenerateSM90_TensorOp_16b_WGMMA_alignx_gemm not generate C.element = DataType.void?
? - Needs Triage
question
Question
#2144
opened Mar 4, 2025 by
henrylhtsang
Previous Next
ProTip!
Exclude everything labeled
bug
with -label:bug.