RIAC as well
License unknown
Use it
Benchmark submission only · no public source · license unknown
No install recipe recorded for this revision.
interfaceunknown
Compatibility
measured onNVIDIA B200
declared hardwareNVIDIA B200
architecturessm_100
dtypesbf16, fp32
Benchmark evidence
1 measurement across 1 GPU, fastest first.
Operation / workload
Hardware
Latency
Rank
Observed
Fused attention QK matmul scale mask softmax backwardsuite of 16 cases
NVIDIA B200
138.5µs
#2 of 27
2026-08-19
Reported · How evidence levels are derived →
Source and license
sourcenot public
revision digestsha256:e5a5e147ee2fc8f01466ecb67ae1c7cce8edf8a2fe16a1fc7ff7000cd6914239
license declaredunknown
license concludedunknown
authorsRIAC as well
imported2026-08-24
Best evidence level for this revision: reported
JSON