Issues 共 1033
[BUG] [CuTeDSL] cute.coalesce doesn't do anything on dynamic shapes
#3469 · lw · 11 天前
[CuTe DSL] setmaxregister_* can be silently dropped by ptxas (C7508) — nothing surfaces to Python; min_blocks_per_mp=1 prevents it
#3406 · zkyue · 2026-07-24
[BUG] CUTLASS DSL
#3395 · yunweili3 · 2026-07-19
[BUG] `CUTE_DSL_KEEP=sass` is ignored in cutlass-dsl 4.6.0
#3365 · tridao · 2026-07-03
[QST] Difference between Fp4 GEMM and Fp4 Ultra Gemm?
#3358 · ankutalev · 2026-06-30
[BUG]
#3327 · harshvardhanagg · 2026-06-17
[BUG] `cute.gemm` rejects valid SM90 WGMMA m64n8k32 in SS mode — degenerate partition layout `(1,1):(0,0)`
#3290 · 16bit-ykiko · 2026-06-02
[QST] Using 2xSM load in MMA kernels with input transform stage
#3284 · infinitron · 2026-05-29
[BUG] [CuTeDSL] `storage.<field>.get_tensor(...)` inside a dynamic `if` block fails with "encountered a user-defined Python object"
#3268 · cheshire · 2026-05-24
[BUG] TMA cache policy requires materialized Int64
#3252 · gau-nernst · 2026-05-20