Struct CutlassNvfp4Gemm 

pub struct CutlassNvfp4Gemm {
    pub tile_m: usize,
    pub tile_n: usize,
    pub tile_k: usize,
    pub cluster_m: usize,
    pub cluster_n: usize,
}

A CUTLASS Blackwell (sm_100) NVFP4 block-scaled GEMM config — the GPU MODE nvfp4_gemm leaderboard target. Operands are FP4 (e2m1) with per-16-element FP8 (ue4m3) block scale factors; accumulate FP32; output FP16. cuBLAS/torch._scaled_mm’s own NVFP4 path is a CUTLASS kernel (as with int8), so our compiler emits the same CollectiveBuilder GEMM and then pulls ahead by consuming the pre-permuted GPU scale factors directly (the reference re-derives them from a CPU tensor inside the timed region — free speedup) + epilogue fusion. NOTE: sm_100 / NVFP4 requires CUTLASS >= 3.7 and a Blackwell GPU; validate on the leaderboard runner (popcorn submit --mode test --gpu B200), not on sm_86 hardware.

Fields

tile_m: usize

tile_n: usize

tile_k: usize

cluster_m: usize

cluster_n: usize

Implementations

impl CutlassNvfp4Gemm

pub fn b200_default() -> Self

A Blackwell NVFP4 config sized to the leaderboard’s speed-of-light shapes (K a multiple of 256).

Trait Implementations

impl Clone for CutlassNvfp4Gemm

fn clone(&self) -> CutlassNvfp4Gemm

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for CutlassNvfp4Gemm

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl PartialEq for CutlassNvfp4Gemm

fn eq(&self, other: &CutlassNvfp4Gemm) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl Copy for CutlassNvfp4Gemm

impl Eq for CutlassNvfp4Gemm

impl StructuralPartialEq for CutlassNvfp4Gemm