FP8 Mamba2 out projection
FP8-quantized output projection in Mamba2 mixer that projects from intermediate size (2688) back to hidden dimension (2688). Uses BlockWise1x128 scaling for activations and BlockWise128x128 scaling for weights.
Reported evidence · last observed 2026-08-22. Reported by source; not independently reproduced. No public source. License unknown.
Current records
Workloadsuite of 16 cases · mean latencyk = 2688 · n = 2688 · block = 128 · k_blocks = 21 · n_blocks = 2116 cases
Not measured on H100 for this workload. Challenges →
127.2µs81% SOL1.00×Reported · license unknown · no source2026-08-22
Measured exactly what you asked. Reported by source; not independently reproduced.
227.6µs81% SOL1.02×Reported · license unknown · no source2026-07-13
Measured exactly what you asked. Reported by source; not independently reproduced.
327.9µs80% SOL1.02×Reported · license unknown · no source2026-07-18
Measured exactly what you asked. Reported by source; not independently reproduced.
431.5µs77% SOL1.16×Reported · license unknown · no source2026-08-02
Measured exactly what you asked. Reported by source; not independently reproduced.
531.6µs77% SOL1.16×Reported · license unknown · no source2026-08-15
Measured exactly what you asked. Reported by source; not independently reproduced.
633.2µs76% SOL1.22×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
734.7µs75% SOL1.27×Reported · license unknown · no source2026-08-03
Measured exactly what you asked. Reported by source; not independently reproduced.
835.0µs74% SOL1.29×Reported · license unknown · no source2026-07-17
Measured exactly what you asked. Reported by source; not independently reproduced.
935.7µs73% SOL1.31×Reported · license unknown · no source2026-08-23
Measured exactly what you asked. Reported by source; not independently reproduced.
1035.8µs73% SOL1.32×Reported · license unknown · no source2026-08-11
Measured exactly what you asked. Reported by source; not independently reproduced.
1136.2µs73% SOL1.33×Reported · license unknown · no source2026-08-20
Measured exactly what you asked. Reported by source; not independently reproduced.
1237.3µs72% SOL1.37×Reported · license unknown · no source2026-08-21
Measured exactly what you asked. Reported by source; not independently reproduced.
1337.3µs72% SOL1.37×Reported · license unknown · no source2026-08-22
Measured exactly what you asked. Reported by source; not independently reproduced.
1437.5µs72% SOL1.38×Reported · license unknown · no source2026-08-19
Measured exactly what you asked. Reported by source; not independently reproduced.
1537.6µs72% SOL1.38×Reported · license unknown · no source2026-08-10
Measured exactly what you asked. Reported by source; not independently reproduced.
1637.6µs72% SOL1.38×Reported · license unknown · no source2026-08-19
Measured exactly what you asked. Reported by source; not independently reproduced.
1737.7µs72% SOL1.39×Reported · license unknown · no source2026-08-18
Measured exactly what you asked. Reported by source; not independently reproduced.
1837.9µs72% SOL1.39×Reported · license unknown · no source2026-08-16
Measured exactly what you asked. Reported by source; not independently reproduced.
1937.9µs72% SOL1.39×Reported · license unknown · no source2026-08-19
Measured exactly what you asked. Reported by source; not independently reproduced.
2038.0µs72% SOL1.40×Reported · license unknown · no source2026-08-10
Measured exactly what you asked. Reported by source; not independently reproduced.
2138.1µs72% SOL1.40×Reported · license unknown · no source2026-08-21
Measured exactly what you asked. Reported by source; not independently reproduced.
2238.1µs72% SOL1.40×Reported · license unknown · no source2026-08-15
Measured exactly what you asked. Reported by source; not independently reproduced.
2338.1µs72% SOL1.40×Reported · license unknown · no source2026-08-17
Measured exactly what you asked. Reported by source; not independently reproduced.
2438.2µs72% SOL1.40×Reported · license unknown · no source2026-08-19
Measured exactly what you asked. Reported by source; not independently reproduced.
2538.4µs71% SOL1.41×Reported · license unknown · no source2026-08-15
Measured exactly what you asked. Reported by source; not independently reproduced.
2638.5µs71% SOL1.42×Reported · license unknown · no source2026-08-24
Measured exactly what you asked. Reported by source; not independently reproduced.
2738.6µs71% SOL1.42×Reported · license unknown · no source2026-08-09
Measured exactly what you asked. Reported by source; not independently reproduced.
2839.1µs71% SOL1.44×Reported · license unknown · no source2026-07-13
Measured exactly what you asked. Reported by source; not independently reproduced.
2939.2µs71% SOL1.44×Reported · license unknown · no source2026-08-09
Measured exactly what you asked. Reported by source; not independently reproduced.
3039.4µs71% SOL1.45×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
3139.6µs71% SOL1.45×Reported · license unknown · no source2026-08-03
Measured exactly what you asked. Reported by source; not independently reproduced.
3239.7µs71% SOL1.46×Reported · license unknown · no source2026-08-09
Measured exactly what you asked. Reported by source; not independently reproduced.
3341.0µs70% SOL1.51×Reported · license unknown · no source2026-07-14
Measured exactly what you asked. Reported by source; not independently reproduced.
3456.6µs62% SOL2.08×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
3564.6µs58% SOL2.38×Reported · license unknown · no source2026-08-11
Measured exactly what you asked. Reported by source; not independently reproduced.
3670.0µs56% SOL2.57×Reported · license unknown · no source2026-08-21
Measured exactly what you asked. Reported by source; not independently reproduced.
3770.2µs56% SOL2.58×Reported · license unknown · no source2026-08-10
Measured exactly what you asked. Reported by source; not independently reproduced.
3893.0µs48% SOL3.42×Reported · license unknown · no source2026-08-22
Measured exactly what you asked. Reported by source; not independently reproduced.
3998.4µs47% SOL3.62×Reported · license unknown · no source2026-07-13
Measured exactly what you asked. Reported by source; not independently reproduced.
40117.8µs42% SOL4.33×Reported · license unknown · no source2026-07-31
Measured exactly what you asked. Reported by source; not independently reproduced.
41123.4µs41% SOL4.54×Reported · license unknown · no source2026-07-26
Measured exactly what you asked. Reported by source; not independently reproduced.
42124.0µs40% SOL4.56×Reported · license unknown · no source2026-07-26
Measured exactly what you asked. Reported by source; not independently reproduced.
43124.1µs40% SOL4.56×Reported · license unknown · no source2026-07-27
Measured exactly what you asked. Reported by source; not independently reproduced.
44929.4µs8% SOL34.18×Reported · license unknown · no source2026-07-13
Measured exactly what you asked. Reported by source; not independently reproduced.