NVIDIA/cutlass
CUDA Templates and Python DSLs for High-Performance Linear Algebra
Is NVIDIA/cutlass beginner-friendly?
In the last 30 days, NVIDIA/cutlass merged 2 of 68 pull requests from outside contributors, and its maintainers usually reply within 4 days. 35 beginner-friendly issues are open now.
- Stars
- 10.7k
- Forks
- 2.2k
- Open beginner issues
- 35
- Indexed issues
- 514
- Avg merge
- 3d 20h
- Merged PRs (30d)
- 4
- Dominant language
- C++
- License
- No license data
- Last GitHub push
- Oct 28, 2026
- Latest indexed
- Sep 19, 2026
- Contributing guide
- No contributing guide
- Code of conduct
- No code of conduct
- Beginner labels
- help wanted good first issue
How to contribute to NVIDIA/cutlass
- There is no contributing guide, so start from the README and a few recently merged pull requests to see how changes are proposed.
- Pick one of the 35 open beginner-friendly issues below and comment that you want to work on it before you start.
Issues somebody may already be working on are listed last. List everything by date
-
bug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 35/100
NVIDIA/cutlass#3797 · 1 comment · 2 reactions ·
Maintainers usually reply within 4 days
-
? - Needs Triage question
Difficulty 2/5 1-3 hours Newbie friendliness 70/100
NVIDIA/cutlass#3796 · 1 comment · 1 reaction ·
Maintainers usually reply within 4 days
-
bug CuTe DSL
Difficulty 4/5 1-2 days Newbie friendliness 48/100
NVIDIA/cutlass#3794 · 1 comment · 1 reaction ·
Maintainers usually reply within 4 days
-
CuTe DSL feature request
Difficulty 4/5 3-5 days Newbie friendliness 35/100
NVIDIA/cutlass#3790 · 1 comment ·
Maintainers usually reply within 4 days
-
CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 48/100
NVIDIA/cutlass#3789 · 1 comment · 4 reactions ·
Maintainers usually reply within 4 days
-
Difficulty 5/5 Over a week Newbie friendliness 35/100
NVIDIA/cutlass#3773 · 1 comment · 1 reaction ·
Maintainers usually reply within 4 days
-
[BUG] testOpenbug CuTe DSL
Difficulty 1/5 Under an hour Newbie friendliness 15/100
Maintainers usually reply within 4 days
-
[QST] SM120 FP8 MMA with FP32 accumulate, plain form against block-scaled form with unit scalesOpen
Difficulty 4/5 3-5 days Newbie friendliness 43/100
NVIDIA/cutlass#3732 · 6 comments ·
Maintainers usually reply within 4 days
-
[BUG] CuTe DSL `CopyUniversalOp` ignores `l1c_evict_priority` again (regression of #3199 since 4.6.0, loads and stores)May be free again A pull request for this issue was closed without being merged. Open? - Needs Triage bug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 25/100
NVIDIA/cutlass#3711 · 2 comments · 1 reaction ·
Maintainers usually reply within 4 days
-
? - Needs Triage bug CuTe DSL
Difficulty 5/5 Over a week Newbie friendliness 38/100
NVIDIA/cutlass#3707 · 7 comments ·
Maintainers usually reply within 4 days
-
Difficulty 5/5 Over a week Newbie friendliness 42/100
NVIDIA/cutlass#3701 · 2 comments ·
Maintainers usually reply within 4 days
-
[BUG] CuTeDSL 4.8.0 nvpu-optimization fails without a diagnostic on a sum of 10 dynamic extentsOpen? - Needs Triage bug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 42/100
NVIDIA/cutlass#3699 · 1 comment · 1 reaction ·
Maintainers usually reply within 4 days
-
? - Needs Triage bug CuTe DSL inactive-30d
Difficulty 5/5 Over a week Newbie friendliness 25/100
NVIDIA/cutlass#3673 · 2 comments ·
Maintainers usually reply within 4 days
-
CuTe DSL inactive-30d
Difficulty 4/5 3-5 days Newbie friendliness 55/100
NVIDIA/cutlass#3672 · 2 comments ·
Maintainers usually reply within 4 days
-
CUTLASS C++ inactive-30d
Difficulty 5/5 Over a week Newbie friendliness 35/100
NVIDIA/cutlass#3606 · 3 comments ·
Maintainers usually reply within 4 days
-
[BUG] 02_dump_reg_shmem faults on sm_120: Fragment is 2-byte aligned but accessed through a 16-byte-aligned pointerMay be free again A pull request for this issue was closed without being merged. Open? - Needs Triage bug CUTLASS C++
Difficulty 4/5 3-5 days Newbie friendliness 55/100
NVIDIA/cutlass#3581 · 6 comments ·
Maintainers usually reply within 4 days
-
CuTe DSL
Difficulty 5/5 Over a week Newbie friendliness 35/100
NVIDIA/cutlass#3573 · 3 comments ·
Maintainers usually reply within 4 days
-
CUTLASS C++
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
NVIDIA/cutlass#3532 · 3 comments ·
Maintainers usually reply within 4 days
-
? - Needs Triage bug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 35/100
NVIDIA/cutlass#3527 · 1 comment · 3 reactions ·
Maintainers usually reply within 4 days
-
CUTLASS C++
Difficulty 3/5 1-2 days Newbie friendliness 76/100
NVIDIA/cutlass#3518 · 1 comment ·
Maintainers usually reply within 4 days
-
CUTLASS C++ duplicate
Difficulty 4/5 3-5 days Newbie friendliness 64/100
NVIDIA/cutlass#3485 · 4 comments ·
Maintainers usually reply within 4 days
-
[Bug][CuTeDSL][Fuzz] `utils.gemm.sm100.epilogue` releases the accumulator before `tcgen05.ld` completes: no `tcgen05.wait::ld` before the arrive that lets the next MMA overwrite TMEMPossibly taken A pull request linked to this issue is open or already merged. Open
Difficulty 2/5 1-3 hours Newbie friendliness 78/100
Maintainers usually reply within 4 days
-
[BUG][CuTeDSL] Blackwell dense GEMM returns wrong results silently for tf32 + 2-CTA + MN-major B with MMA tiler N 32 / 96 / 160 / 224: get_smem_layout_atom_ab picks MN_SW64 for a 32-bit MN-major operandPossibly taken @yunweili3 claimed this 5 days ago. Open
Difficulty 4/5 3-5 days Newbie friendliness 25/100
NVIDIA/cutlass#3792 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG][CuTeDSL] Blackwell dense_gemm example aborts (SIGABRT in cute.slice_) for tf32 + 2-CTA + B K-major with MMA tiler N 96 / 160, which can_implement() acceptsPossibly taken @yunweili3 claimed this 7 days ago. Open
Difficulty 4/5 3-5 days Newbie friendliness 25/100
NVIDIA/cutlass#3784 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] TMA epilogue hangs with a null beta_ptr, nonzero dBeta and nonzero scalar beta (load warp and epilogue disagree on loading C)Possibly taken @RohanQ17 claimed this 1 day ago. OpenCUTLASS C++
Difficulty 4/5 3-5 days Newbie friendliness 30/100
NVIDIA/cutlass#3783 · 2 comments ·
Maintainers usually reply within 4 days
-
[BUG] SM100 mixed TMA/cp.async GEMM: scheduler (CLC) pipeline does not count the epilogue-load warp, hangs when C is loadedPossibly taken @builtin-pb claimed this 7 days ago. OpenCUTLASS C++
Difficulty 4/5 3-5 days Newbie friendliness 30/100
Maintainers usually reply within 4 days
-
[BUG] Example 77 MLA fused split-KV reduction (--fuse_reduction) can hang or return wrong outputPossibly taken @builtin-pb claimed this 11 days ago. OpenCUTLASS C++
Difficulty 4/5 3-5 days Newbie friendliness 38/100
Maintainers usually reply within 4 days
-
[DOC] IKET quick start omits the GEMM GPU requirement and references a missing examplePossibly taken @0z5a claimed this 11 days ago. Open
Difficulty 1/5 Under an hour Newbie friendliness 82/100
NVIDIA/cutlass#3761 · 1 comment ·
Maintainers usually reply within 4 days
-
[DOC] Dynamic-layout guide example uses incompatible singleton stridesPossibly taken @0z5a claimed this 11 days ago. Open
Difficulty 2/5 1-3 hours Newbie friendliness 25/100
NVIDIA/cutlass#3760 · 1 comment ·
Maintainers usually reply within 4 days
-
[DOC] TVM FFI guide examples fail due to missing imports, exception type and loader ABIPossibly taken @0z5a claimed this 11 days ago. Open
Difficulty 2/5 1-3 hours Newbie friendliness 35/100
NVIDIA/cutlass#3759 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] CuTe DSL: SIGSEGV compiling integer ** (int32/int64), including 2 ** 3Possibly taken @yunweili3 claimed this 18 days ago. Openbug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 38/100
NVIDIA/cutlass#3735 · 3 comments ·
Maintainers usually reply within 4 days
-
[BUG] `cute.ceil_div` throws internal error with int64 inputPossibly taken @yunweili3 claimed this 19 days ago. Open? - Needs Triage bug CuTe DSL
Difficulty 3/5 1-2 days Newbie friendliness 55/100
NVIDIA/cutlass#3734 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] Cache policy from `cute.copy` is dropped when the copy atom is created with `invariant=True`Possibly taken @yunweili3 claimed this 22 days ago. Open? - Needs Triage bug CuTe DSL
Difficulty 4/5 1-2 days Newbie friendliness 28/100
NVIDIA/cutlass#3726 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] SM90 TMA GEMM mainloops rely on elect_one_sync() electing lane 0 (TMA issuer vs. PipelineTmaAsync is_leader)Possibly taken @yunweili3 claimed this 23 days ago. OpenCUTLASS C++
Difficulty 4/5 3-5 days Newbie friendliness 48/100
NVIDIA/cutlass#3722 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] SM90 cp.async + GMMA collectives: no fence.proxy.async between cp.async operand writes and wgmma readsPossibly taken @yunweili3 claimed this 23 days ago. OpenCUTLASS C++
Difficulty 4/5 3-5 days Newbie friendliness 42/100
NVIDIA/cutlass#3721 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] Grouped GEMM and grouped Rank2K kernels build operand iterators from a temporary ParamsPossibly taken @BwL1289 claimed this 24 days ago. OpenCUTLASS C++
Difficulty 4/5 3-5 days Newbie friendliness 55/100
Maintainers usually reply within 4 days
-
[BUG] CuTe DSL from_dlpack rejects float8_e4m3fn tensors (dtype_map lists Float8E5M2 only)Possibly taken @Ammar-Alnagar claimed this 20 days ago. Open
Difficulty 2/5 1-3 hours Newbie friendliness 74/100
NVIDIA/cutlass#3716 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] CuTe DSL >=4.6.0: `export_to_c` does not apply explicit `preferred_smem_carveout`Possibly taken @DQR-Happy claimed this 21 days ago. Open? - Needs Triage bug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 53/100
NVIDIA/cutlass#3715 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] [BUG] `cute.arch.store` doesn't work with fp8 vectorPossibly taken @yunweili3 claimed this 26 days ago. Open? - Needs Triage bug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 25/100
NVIDIA/cutlass#3712 · 1 comment · 1 reaction ·
Maintainers usually reply within 4 days
-
[BUG] mma_atom_call for mma.sync is eliminated if C is RZPossibly taken @yunweili3 claimed this 35 days ago. Open? - Needs Triage bug CuTe DSL inactive-30d
Difficulty 4/5 3-5 days Newbie friendliness 55/100
NVIDIA/cutlass#3689 · 3 comments ·
Maintainers usually reply within 4 days
-
[BUG] CuTe DSL: `a % b` gives different results for dynamic and constant operands when the signs differPossibly taken @yunweili3 claimed this 35 days ago. Openbug CuTe DSL inactive-30d
Difficulty 3/5 1-2 days Newbie friendliness 76/100
NVIDIA/cutlass#3687 · 2 comments ·
Maintainers usually reply within 4 days
-
[BUG] CuTe DSL: cute.compile stalls for minutes when no GPU is visiblePossibly taken @Hashim1999164 claimed this 35 days ago. OpenCuTe DSL inactive-30d
Difficulty 3/5 1-2 days Newbie friendliness 72/100
NVIDIA/cutlass#3681 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] CuTe DSL: export_to_c fails for a --gpu-arch other than the current GPU'sPossibly taken @SammyTourani claimed this 35 days ago. Openbug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 54/100
NVIDIA/cutlass#3680 · 2 comments ·
Maintainers usually reply within 4 days
-
[BUG][CuTeDSL] cute.copy(..., tma_desc_ptr=p) with a gmem-space p passes the Python checks, then fails in libNVVMPossibly taken @yunweili3 claimed this 37 days ago. OpenCuTe DSL inactive-30d
Difficulty 4/5 3-5 days Newbie friendliness 68/100
NVIDIA/cutlass#3671 · 2 comments ·
Maintainers usually reply within 4 days
-
[BUG][CuTeDSL] `cute.copy(..., tma_desc_ptr=TensorMap.get_ptr())` is silently ignored on tiled TMA loadsPossibly taken @yunweili3 claimed this 37 days ago. OpenCuTe DSL inactive-30d
Difficulty 4/5 3-5 days Newbie friendliness 65/100
NVIDIA/cutlass#3670 · 2 comments ·
Maintainers usually reply within 4 days
-
[BUG][CuTe DSL] PipelineTmaUmma arrival masks include an out-of-cluster "v-peer" bit when the cluster's v-extent is 1, causing CGA out-of-range faults (Xid 13)Possibly taken @yunweili3 claimed this 39 days ago. Openinactive-30d
Difficulty 3/5 1-2 days Newbie friendliness 40/100
NVIDIA/cutlass#3658 · 2 comments ·
Maintainers usually reply within 4 days
-
[BUG] Instantiated tile sizes for sm120 fail to meet kernel's layout assertion during cutlass_profiler compilePossibly taken @LiRunGuo claimed this 39 days ago. Open? - Needs Triage bug CUTLASS C++ inactive-30d
Difficulty 4/5 3-5 days Newbie friendliness 50/100
NVIDIA/cutlass#3646 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] Incorrect results in 2CTA MMA example (4.8.0dev0) due to barrier scope & missing SFB waitPossibly taken @yunweili3 claimed this 41 days ago. Open? - Needs Triage bug CuTe DSL
Difficulty 4/5 3-5 days Newbie friendliness 68/100
NVIDIA/cutlass#3643 · 1 comment ·
Maintainers usually reply within 4 days
-
[BUG] Operand initialized with a constant doesn't work with `inline_ptx` as an operandPossibly taken @0z5a claimed this 43 days ago. Open? - Needs Triage bug CuTe DSL inactive-30d
Difficulty 4/5 3-5 days Newbie friendliness 52/100
NVIDIA/cutlass#3640 · 3 comments ·
Maintainers usually reply within 4 days
-
[BUG] Constexpr missing in dynamic control-flow regionsPossibly taken @nodeeeeee claimed this 58 days ago. Open? - Needs Triage bug CuTe DSL inactive-30d
Difficulty 3/5 1-2 days Newbie friendliness 68/100
NVIDIA/cutlass#3632 · 1 comment ·
Maintainers usually reply within 4 days
Showing the newest 100
This page lists what was indexed most recently. Advanced filter has the whole inventory, narrowed by language, difficulty and how long a task takes.