Struct MoeCutlassHostTables 

pub struct MoeCutlassHostTables {
    pub gate_packed: Vec<u64>,
    pub gate_sfb: Vec<u64>,
    pub gate_scale2: Vec<f32>,
    pub up_packed: Vec<u64>,
    pub up_sfb: Vec<u64>,
    pub up_scale2: Vec<f32>,
    pub down: Option<MoeCutlassDownHostTables>,
}

2026-09-25: Host copies of the per-expert pointer and scale tables for the CUTLASS grouped path, which takes them on the host. MoeLayer::build_cutlass_grouped_sfb builds them at load into the layer’s cutlass_grouped_host, so they are dropped with the layer.

Fields

gate_packed: Vec<u64>

gate_sfb: Vec<u64>

gate_scale2: Vec<f32>

up_packed: Vec<u64>

up_sfb: Vec<u64>

up_scale2: Vec<f32>

down: Option<MoeCutlassDownHostTables>

2026-09-25: None when the layer has no down-projection scale table; the grouped down branch of forward_prefill_routed is then not taken.

Implementations

impl MoeCutlassHostTables

pub fn snapshot( gpu: &dyn GpuBackend, num_experts: usize, gate_packed: DevicePtr, gate_sfb: Vec<u64>, gate_scale2: DevicePtr, up_packed: DevicePtr, up_sfb: Vec<u64>, up_scale2: DevicePtr, down: Option<(DevicePtr, Vec<u64>, DevicePtr)>, ) -> Result<Self>

2026-09-25: Snapshot the grouped-path tables. The SFB pointer vectors are taken by value because build_cutlass_grouped_sfb builds them on the host; the packed and scale2 tables are copied from the device here.