pub fn w8a16_gemv_batch16_ncol2_strided(
gpu: &dyn GpuBackend,
kernel: KernelHandle,
input: DevicePtr,
weight: DevicePtr,
block_scale: DevicePtr,
output: DevicePtr,
m: u32,
n: u32,
k: u32,
a_row_stride: u32,
c_row_stride: u32,
stream: u64,
) -> Result<()>Expand description
N_COLS=2, explicit A and C row pitches in ELEMENTS — for the multi-seq
[n, per_seq_qkv] decode buffer, exactly as w8a16_gemv_batch16_strided.
Kernel: w8a16_gemv_batch16_ncol2_strided (module w8a16_gemv_ncol).