pub fn w8a8_act_quant_silu(
gpu: &dyn GpuBackend,
kernels: &W8a8Kernels,
scale: W8a8Scale,
gate: DevicePtr,
up: DevicePtr,
ld: u32,
rows: usize,
k: u32,
scratch: &W8a8Scratch,
stream: u64,
) -> Result<()>Expand description
2026-09-28: w8a8_act_quant of bf16(silu(gate) * up), computed in the
same launch with the arithmetic of moe_silu_mul: gate and up are
[rows, ld] BF16, the first k columns read.