pub fn fused_k_norm_rope_cache_write_fp8(
gpu: &dyn GpuBackend,
kernel: KernelHandle,
k_in: DevicePtr,
k_norm_weight: DevicePtr,
positions: DevicePtr,
k_cache_fp8: DevicePtr,
slot_mapping: DevicePtr,
num_tokens: u32,
num_kv_heads: u32,
head_dim: u32,
rotary_dim: u32,
block_size: u32,
rms_eps: f32,
theta: f32,
inv_scale: f32,
stream: u64,
) -> Result<()>Expand description
FP8-output sibling of fused_k_norm_rope_cache_write_bf16. Same
semantics; one fewer BF16 round before the saturating FP8 cast.