diff --git a/src/llama-quant.cpp b/src/llama-quant.cpp
index fb7982655..c91af4cbd 100644
--- a/src/llama-quant.cpp
+++ b/src/llama-quant.cpp
@@ -514,7 +514,7 @@ static void llama_model_quantize_impl(const std::string & fname_inp, const std::
 
     // mmap consistently increases speed Linux, and also increases speed on Windows with
     // hot cache. It may cause a slowdown on macOS, possibly related to free memory.
-#if defined(__linux__) || defined(_WIN32)
+#if (defined(__linux__) && !defined(__s390x__)) || defined(_WIN32)
     constexpr bool use_mmap = true;
 #else
     constexpr bool use_mmap = false;