Search results

[BUG] UInt64 constants are compared to i64::max instead of u64::max in consteval

? - Needs Triage

bugSomething isn't workingSomething isn't working

CuTe DSL

Status: Open.

#3312 In NVIDIA/cutlass;

· p-adema opened on Jun 11, 2026

[BUG][CuTeDSL] AOT compilation fails with GPUArch("sm_100f") + TVM-FFI on B200 node

? - Needs Triage

bugSomething isn't workingSomething isn't working

CuTe DSL

Status: Open.

#3298 In NVIDIA/cutlass;

· ktaebum opened on Jun 4, 2026

[FEA] Tensor manipulation methods for cute.Tensor

? - Needs Triage

CuTe DSL

feature requestNew feature or requestNew feature or request

Status: Open.

#3296 In NVIDIA/cutlass;

· nandor opened on Jun 4, 2026

[BUG] cute.gemm rejects valid SM90 WGMMA m64n8k32 in SS mode — degenerate partition layout (1,1):(0,0)

? - Needs Triage

bugSomething isn't workingSomething isn't working

CuTe DSL

Status: Open.

#3290 In NVIDIA/cutlass;

· 16bit-ykiko opened on Jun 2, 2026

[QST]Memory Ordering of atomicAdd and tma_reduce_add

? - Needs Triage

questionQuestionQuestion

Status: Open.

#3282 In NVIDIA/cutlass;

· kindwyf opened on May 28, 2026

[BUG] CuTeDSL 4.5.2: MmaFP8Op.make_fragment_A segfaults on SM120 (Blackwell)

? - Needs Triage

bugSomething isn't workingSomething isn't working

CuTe DSL

Status: Open.

#3281 In NVIDIA/cutlass;

· L14nY1Wang opened on May 28, 2026

[BUG] Allocating tensor makes calling cute.jit a lot slower even it doesn't use the allocated tensor

? - Needs Triage

bugSomething isn't workingSomething isn't working

CuTe DSL

Status: Open.

#3275 In NVIDIA/cutlass;

· kainzhong opened on May 26, 2026

[BUG] Potential overflow issue in packed_stride.hpp

Status: Open.

#3269 In NVIDIA/cutlass;

· Jovp opened on May 24, 2026

[CuTeDSL] nvvm.load.ext rejects BFloat16: "Unsupported FP type for ExtLoadOp"

? - Needs Triage

bugSomething isn't workingSomething isn't working

CuTe DSL

Status: Open.

#3266 In NVIDIA/cutlass;

· cheshire opened on May 24, 2026

[Feature Request] SM_120/SM_121 CollectiveBuilder specialization for tensor/token-scaled FP8 grouped GEMM (ptr-array)

Status: Open.

#3263 In NVIDIA/cutlass;

· tgmerritt opened on May 23, 2026

[BUG] cutlass-dsl 4.5.1: -libs-base and -libs-cu13 wheels write the same path

? - Needs Triage

bugSomething isn't workingSomething isn't working

CuTe DSL

Status: Open.

#3259 In NVIDIA/cutlass;

· arpera opened on May 21, 2026

[FEA] Documentation for building nvidia-cutlass-dsl-libs-base package

? - Needs Triage

CUTLASS C++

feature requestNew feature or requestNew feature or request

Status: Open.

#3257 In NVIDIA/cutlass;

· Flamefire opened on May 21, 2026