MoE expert parallel execution with weighted aggregation
Complete MoE expert parallel execution with weighted aggregation. After routing determines expert assignments, this operation performs parallel expert computations across all 16 experts with dynamic token assignment. Each expert processes assigned tokens through a gated MLP (gate_proj, up_proj, SiLU, multiply, down_proj), then results are weighted by routing weights and aggregated back to original token positions via scattered writes.
Reported evidence · last observed 2026-07-10. Reported by source; not independently reproduced. No public source. License unknown.
Current records
Workloadsuite of 16 cases · mean latencyhidden_size = 8192 · num_experts = 16 · intermediate_size = 24576 · num_experts_per_tok = 216 cases
Not measured on H100 for this workload. Challenges →
14.93ms68% SOL1.00×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
24.98ms67% SOL1.01×Reported · license unknown · no source2026-07-21
Measured exactly what you asked. Reported by source; not independently reproduced.
35.00ms67% SOL1.01×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
45.15ms65% SOL1.04×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
55.20ms64% SOL1.05×Reported · license unknown · no source2026-08-14
Measured exactly what you asked. Reported by source; not independently reproduced.
65.25ms63% SOL1.06×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
75.42ms60% SOL1.10×Reported · license unknown · no source2026-08-20
Measured exactly what you asked. Reported by source; not independently reproduced.
85.44ms60% SOL1.10×Reported · license unknown · no source2026-08-19
Measured exactly what you asked. Reported by source; not independently reproduced.
95.44ms60% SOL1.10×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
105.53ms59% SOL1.12×Reported · license unknown · no source2026-07-17
Measured exactly what you asked. Reported by source; not independently reproduced.
115.59ms59% SOL1.13×Reported · license unknown · no source2026-08-08
Measured exactly what you asked. Reported by source; not independently reproduced.
125.77ms56% SOL1.17×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
136.13ms52% SOL1.24×Reported · license unknown · no source2026-07-10
Measured exactly what you asked. Reported by source; not independently reproduced.
148.28ms38% SOL1.68×Reported · license unknown · no source2026-07-27
Measured exactly what you asked. Reported by source; not independently reproduced.