Function moe_expert_silu_down_shared_fp8_batch2 

pub fn moe_expert_silu_down_shared_fp8_batch2(
    gpu: &dyn GpuBackend,
    kernel: KernelHandle,
    gate_out: DevicePtr,
    up_out: DevicePtr,
    dp_w: DevicePtr,
    dp_s: DevicePtr,
    output: DevicePtr,
    indices: DevicePtr,
    sh_gate_in: DevicePtr,
    sh_up_in: DevicePtr,
    sh_down: &Fp8Weight,
    sh_down_out: DevicePtr,
    n: u32,
    k: u32,
    top_k: u32,
    stream: u64,
) -> Result<()>

2026-09-25: FP8 fused SiLU+down GEMV over the routed experts and the shared expert, for two tokens.