hunyuanimage-3-0
0 of 3 operations have a usable best known on NVIDIA B200 · 3 without one
Best known on NVIDIA B200
Fused final layer upsample with adaptive normsuite of 16 casesAmir M. Mir | SF Tensornot usable1.58msReported · license unknown · no source2026-07-26
No entry in this cohort passes the usability policy (no public source, no install recipe, license unknown); the fastest known is shown. Measured by the source's own harness. Ranked by its metric under ranking-v1. 26 more entries in this cohort.
Fused GQA QKV RoPE QK norm backwardsuite of 16 casesRIAC as wellnot usable156.9µsReported · license unknown · no source2026-08-21
No entry in this cohort passes the usability policy (no public source, no install recipe, license unknown); the fastest known is shown. Measured by the source's own harness. Ranked by its metric under ranking-v1. 18 more entries in this cohort.
Batched 2d RoPE position encoding backwardsuite of 16 casesRIAC as wellnot usable4.59µsReported · license unknown · no source2026-08-20
No entry in this cohort passes the usability policy (no public source, no install recipe, license unknown); the fastest known is shown. Measured by the source's own harness. Ranked by its metric under ranking-v1. 16 more entries in this cohort.
Coverage gaps on NVIDIA B200
No eligible evidence yet for these operations. Challenges →