use __nv_fp4x4_e2m1

ggml/src/ggml-cuda/quantize.cu

@@ -122,6 +122,26 @@ static __global__ void quantize_mmq_mxfp4(const float * __restrict__ x,
         const uint8_t e = compute_e8m0_scale(amax);
         scales[b] = e;
         const float inv_s = (amax == 0.0f) ? 0.0f : __frcp_rn(ggml_cuda_e8m0_to_fp32(e));
+
+#if CUDART_VERSION >= 12040
+        // Use hardware FP4 conversion: pre-scale and gather 4 floats, then convert+pack
+        const float scaled_val = xi * inv_s;
+
+        // Gather 4 scaled floats in the order matching __nv_fp4x4_e2m1 packing:
+        // float4(x,y,z,w) -> 16-bit with bits [3:0]=x, [7:4]=y, [11:8]=z, [15:12]=w
+        // This produces byte0 = (y<<4)|x, byte1 = (w<<4)|z
+        const float val0 = __shfl_sync(0xFFFFFFFF, scaled_val, base,      WARP_SIZE);  // -> low nibble byte 0
+        const float val1 = __shfl_sync(0xFFFFFFFF, scaled_val, base + 16, WARP_SIZE);  // -> high nibble byte 0
+        const float val2 = __shfl_sync(0xFFFFFFFF, scaled_val, base + 1,  WARP_SIZE);  // -> low nibble byte 1
+        const float val3 = __shfl_sync(0xFFFFFFFF, scaled_val, base + 17, WARP_SIZE);  // -> high nibble byte 1
+
+        if (lane_in_group == 0) {
+            // Convert 4 floats -> packed 16-bit FP4 in one step
+            __nv_fp4x4_e2m1 fp4_packed(make_float4(val0, val1, val2, val3));
+            yqs2[pair_idx_in_block * 16 + b * 8 + group_id] = reinterpret_cast<const char2&>(fp4_packed);
+        }
+#else
+        // Fallback: manual FP4 conversion using LUT
         const uint8_t q_val = ggml_cuda_float_to_fp4_e2m1(xi, inv_s);
 
         const uint8_t q_lo_0 = __shfl_sync(0xFFFFFFFF, q_val, base,      WARP_SIZE);
@@ -135,6 +155,7 @@ static __global__ void quantize_mmq_mxfp4(const float * __restrict__ x,
             q.y = (q_hi_1 << 4) | q_lo_1;
             yqs2[pair_idx_in_block * 16 + b * 8 + group_id] = q;
         }
+#endif // CUDART_VERSION >= 12040
     }
 
     if (lane_id_32 == 0) {

ggml/src/ggml-cuda/vendors/cuda.h

@@ -10,6 +10,10 @@
 #include <cuda_fp8.h>
 #endif // CUDART_VERSION >= 12050
 
+#if CUDART_VERSION >= 12040
+#include <cuda_fp4.h>
+#endif // CUDART_VERSION >= 12040
+
 #if CUDART_VERSION < 11020
 #define CU_DEVICE_ATTRIBUTE_VIRTUAL_MEMORY_MANAGEMENT_SUPPORTED CU_DEVICE_ATTRIBUTE_VIRTUAL_ADDRESS_MANAGEMENT_SUPPORTED
 #define CUBLAS_TF32_TENSOR_OP_MATH CUBLAS_TENSOR_OP_MATH