GEMM n8192 k3072
0 eligible runs
gemm
General matrix multiply (GEMM) C = A @ B.T. Captured from MiniMax M2 attn.qkv_proj (fused q+k+v: 48*128 + 2*8*128 = 8192).
Current records
No published measurement for the selected workload.
Implementations
Implementation
Runtime
Best latency
Evidence
Availability
Semantics
Inputs and outputs
afp16 [m, k]
bfp16 [n, k]
cfp16 [m, n]
Axes and behavior
kconstant = 3072
mvariable
nconstant = 8192
determinismunspecified
constraintsNo mutation or aliasing
Identity
No source imports for this operation yet.How records are decidedJSON