Function prefill_attention_paged_fp8_batched_64 

pub fn prefill_attention_paged_fp8_batched_64(
    gpu: &dyn GpuBackend,
    kernel: KernelHandle,
    q: DevicePtr,
    k_cache: DevicePtr,
    v_cache: DevicePtr,
    output: DevicePtr,
    block_table_ptrs: DevicePtr,
    batch_size: u32,
    cu_seqlens: DevicePtr,
    kv_lens: DevicePtr,
    q_len: u32,
    kv_len: u32,
    q_offset: u32,
    num_q_heads: u32,
    num_kv_heads: u32,
    head_dim: u32,
    cache_block_size: u32,
    sliding_window: u32,
    inv_sqrt_d: f32,
    k_scale: f32,
    v_scale: f32,
    cache_stride: u64,
    stream: u64,
) -> Result<()>

Batched FP8-KV paged prefill attention (BR=64, 256-thread variant).