diff --git a/Makefile b/Makefile
index 9dd8a51c4..dbf9d52b7 100644
--- a/Makefile
+++ b/Makefile
@@ -171,11 +171,11 @@ ifdef LLAMA_CUDA_DMMV_Y
 else
 	NVCCFLAGS += -DGGML_CUDA_DMMV_Y=1
 endif # LLAMA_CUDA_DMMV_Y
-#ifdef LLAMA_CUDA_KQUANTS_ITER
+ifdef LLAMA_CUDA_KQUANTS_ITER
 	NVCCFLAGS += -DK_QUANTS_PER_ITERATION=$(LLAMA_CUDA_KQUANTS_ITER)
-#else
+else
 	NVCCFLAGS += -DK_QUANTS_PER_ITERATION=2
-#endif
+endif
 ggml-cuda.o: ggml-cuda.cu ggml-cuda.h
 	$(NVCC) $(NVCCFLAGS) $(CXXFLAGS) -Wno-pedantic -c $< -o $@
 endif # LLAMA_CUBLAS
diff --git a/ggml-cuda.cu b/ggml-cuda.cu
index 46f50e492..91c3a3b70 100644
--- a/ggml-cuda.cu
+++ b/ggml-cuda.cu
@@ -169,6 +169,8 @@ static_assert(sizeof(block_q6_K) == sizeof(ggml_fp16_t) + 13*QK_K/16, "wrong q6_
 
 #ifndef K_QUANTS_PER_ITERATION
 #define K_QUANTS_PER_ITERATION 2
+#else
+static_assert(K_QUANTS_PER_ITERATION == 1 || K_QUANTS_PER_ITERATION == 2, "K_QUANTS_PER_ITERATION must be 1 or 2");
 #endif
 
 static __global__ void add_f32(const float * x, const float * y, float * dst, const int k) {