FFN GELU projection fused backward
Backward pass for fused FFN block with GELU activation, residual connection, and layer normalization. Computes gradients for hidden_states, fc1_weight, fc1_bias, fc2_weight, fc2_bias, ln_weight, and ln_bias.
Reported evidence · last observed 2026-08-18. Reported by source; not independently reproduced. No public source. License unknown.
Current records
Workloadsuite of 16 cases · mean latencyhidden_size = 512 · intermediate_size = 204816 cases
Not measured on H100 for this workload. Challenges →
1115.1µs75% SOL1.00×Reported · license unknown · no source2026-08-18
Measured exactly what you asked. Reported by source; not independently reproduced.
2120.3µs74% SOL1.04×Reported · license unknown · no source2026-08-20
Measured exactly what you asked. Reported by source; not independently reproduced.
3133.8µs71% SOL1.16×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
4149.4µs68% SOL1.30×Reported · license unknown · no source2026-07-19
Measured exactly what you asked. Reported by source; not independently reproduced.
5149.9µs68% SOL1.30×Reported · license unknown · no source2026-08-15
Measured exactly what you asked. Reported by source; not independently reproduced.
6150.2µs68% SOL1.30×Reported · license unknown · no source2026-08-15
Measured exactly what you asked. Reported by source; not independently reproduced.
7150.5µs68% SOL1.31×Reported · license unknown · no source2026-08-15
Measured exactly what you asked. Reported by source; not independently reproduced.
8157.5µs68% SOL1.37×Reported · license unknown · no source2026-07-17
Measured exactly what you asked. Reported by source; not independently reproduced.
9158.6µs67% SOL1.38×Reported · license unknown · no source2026-07-24
Measured exactly what you asked. Reported by source; not independently reproduced.
10162.3µs67% SOL1.41×Reported · license unknown · no source2026-07-14
Measured exactly what you asked. Reported by source; not independently reproduced.
11183.3µs64% SOL1.59×Reported · license unknown · no source2026-08-24
Measured exactly what you asked. Reported by source; not independently reproduced.
12184.5µs63% SOL1.60×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
13187.7µs63% SOL1.63×Reported · license unknown · no source2026-08-15
Measured exactly what you asked. Reported by source; not independently reproduced.
14228.0µs58% SOL1.98×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
15236.0µs57% SOL2.05×Reported · license unknown · no source2026-08-13
Measured exactly what you asked. Reported by source; not independently reproduced.
16258.7µs54% SOL2.25×Reported · license unknown · no source2026-08-11
Measured exactly what you asked. Reported by source; not independently reproduced.
17263.8µs54% SOL2.29×Reported · license unknown · no source2026-08-22
Measured exactly what you asked. Reported by source; not independently reproduced.
18347.6µs46% SOL3.02×Reported · license unknown · no source2026-08-07
Measured exactly what you asked. Reported by source; not independently reproduced.
19360.5µs45% SOL3.13×Reported · license unknown · no source2026-08-10
Measured exactly what you asked. Reported by source; not independently reproduced.
20436.7µs40% SOL3.79×Reported · license unknown · no source2026-07-23
Measured exactly what you asked. Reported by source; not independently reproduced.
21577.8µs33% SOL5.02×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
22769.3µs27% SOL6.68×Reported · license unknown · no source2026-08-10
Measured exactly what you asked. Reported by source; not independently reproduced.
231.56ms19% SOL13.54×Reported · license unknown · no source2026-08-02
Measured exactly what you asked. Reported by source; not independently reproduced.