diff --git a/ggml/src/ggml-cuda/mmq.cuh b/ggml/src/ggml-cuda/mmq.cuh
index 9618579424..209555ce39 100644
--- a/ggml/src/ggml-cuda/mmq.cuh
+++ b/ggml/src/ggml-cuda/mmq.cuh
@@ -384,7 +384,8 @@ static __device__ __forceinline__ void vec_dot_q4_0_q8_1_dp4a(
                 int u[2*VDR_Q4_0_Q8_1_MMQ];
 
                 constexpr int max_cpy = ggml_cuda_get_max_cpy_bytes();
-                constexpr int mcpy_int = max_cpy / sizeof(int);  
+                constexpr int mcpy_int = max_cpy / sizeof(int);
+                static_assert(VDR_Q4_0_Q8_1_MMQ == 4, "bad VDR_Q4_0_Q8_1_MMQ");
 
                 int tmp0[4], tmp1[4];