FP8 Mamba2 SSM discretization
FP8-quantized SSM parameter discretization in Mamba2 that computes dt (time step), dA (discretized A matrix), and dB (discretized B matrix) using FP8 linear projections with blockwise scaling.
Reported evidence · last observed 2026-07-27. Reported by source; not independently reproduced. No public source. License unknown.
Current records
Not measured on H100 for this workload. Challenges →
116.3µs79% SOL1.00×Reported · license unknown · no source2026-07-27
Measured exactly what you asked. Reported by source; not independently reproduced.
216.5µs79% SOL1.01×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
316.5µs79% SOL1.01×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
416.5µs79% SOL1.01×Reported · license unknown · no source2026-08-01
Measured exactly what you asked. Reported by source; not independently reproduced.
516.6µs79% SOL1.01×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
616.6µs79% SOL1.01×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
717.8µs78% SOL1.09×Reported · license unknown · no source2026-08-04
Measured exactly what you asked. Reported by source; not independently reproduced.
818.5µs77% SOL1.13×Reported · license unknown · no source2026-07-18
Measured exactly what you asked. Reported by source; not independently reproduced.
919.0µs76% SOL1.16×Reported · license unknown · no source2026-08-04
Measured exactly what you asked. Reported by source; not independently reproduced.
1020.2µs75% SOL1.24×Reported · license unknown · no source2026-08-03
Measured exactly what you asked. Reported by source; not independently reproduced.
1120.6µs75% SOL1.26×Reported · license unknown · no source2026-08-02
Measured exactly what you asked. Reported by source; not independently reproduced.
1221.7µs74% SOL1.33×Reported · license unknown · no source2026-08-04
Measured exactly what you asked. Reported by source; not independently reproduced.
1326.9µs69% SOL1.64×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
1430.7µs67% SOL1.88×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
1540.8µs60% SOL2.50×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
1673.0µs46% SOL4.47×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.