pub fn w8a8_proj(
gpu: &dyn GpuBackend,
kernels: &W8a8Kernels,
w: &W8a8Weight,
x: DevicePtr,
ldx: u32,
rows: usize,
out: DevicePtr,
ldc: u32,
scratch: &W8a8Scratch,
stream: u64,
) -> Result<()>Expand description
2026-09-28: One W8A8 projection: quantize x ([rows, ldx] BF16) and
multiply, writing out ([rows, ldc] BF16, the first w.n() columns).