qwen3-4b-instruct-2507
best known per operation on the selected GPU
2 operations · 2 families · 72 eligible runs
0 of 2 operations have a usable best known on NVIDIA B200 · 2 without one
Best known on NVIDIA B200
Operation · workload
Best known
Record
Trust
Observed
rmsnorm1 operation
Fused residual RMS norm backwardsuite of 16 casesRecursivenot usable42.0µsReported · license unknown · no source2026-06-06
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. 19 more entries in this cohort.
no sourcelicense unknownno install recipeCohort on the operation page →Run detail →
rope1 operation
Rotary embedding with attention scaling backwardsuite of 16 casesdoubleAInot usable4.93µsReported · license unknown · no source2026-07-10
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. 17 more entries in this cohort.
no sourcelicense unknownno install recipeCohort on the operation page →Run detail →
Coverage gaps on NVIDIA B200
Operation
Family
Status
rmsnorm
measured · no usable entry (no public source, no install recipe, license unknown)
rope
measured · no usable entry (no public source, no install recipe, license unknown)
No eligible evidence yet for these operations. Challenges →
Sources: NVIDIA SOL-ExecBench (2026-08-24)How comparison worksJSON