Function nvfp4_grouped_gate_up 

pub fn nvfp4_grouped_gate_up(
    a: u64,
    gate_packed_ptrs: &[u64],
    gate_scale_ptrs: &[u64],
    gate_scale2_vals: &[f32],
    up_packed_ptrs: &[u64],
    up_scale_ptrs: &[u64],
    up_scale2_vals: &[f32],
    c_gate: u64,
    c_up: u64,
    expert_offsets: &[i32],
    n: u32,
    k: u32,
    stream: u64,
) -> Result<()>

Grouped (per-expert) NVFP4 fused gate_up GEMM — Holo MoE Phase-1 escape-hatch path. Dispatches the proven Sm120 NVFP4 collective once per active expert over its token slice; bit-faithful to nvfp4_gemm_bf16_act_weight_t (it IS that collective), at one launch per expert. Used to validate that the FP4 math integrates correctly in grouped form before the hand-rolled block-scaled mma (Phase 2).

a is bf16 [M_total, K]; expert e owns rows [expert_offsets[e], expert_offsets[e+1]). *_packed_ptrs/*_scale_ptrs are device-pointer arrays (one per expert) in the pack_bf16_weight_to_nvfp4_t layout ([N,K/2] + [K/16,N]); the *_scale2_vals and expert_offsets slices are HOST arrays.