diff --git a/csrc/cpu/utils.hpp b/csrc/cpu/utils.hpp
index d8399c56f6af8..d3def306b8069 100644
--- a/csrc/cpu/utils.hpp
+++ b/csrc/cpu/utils.hpp
@@ -6,6 +6,10 @@
 #include <cstdint>
 #include <unistd.h>
 
+#if defined(__APPLE__)
+  #include <sys/sysctl.h>
+#endif
+
 #include "cpu_types.hpp"
 
 namespace cpu_utils {
@@ -21,10 +25,12 @@ struct VecTypeTrait<float> {
   using vec_t = vec_op::FP32Vec16;
 };
 
+#if !defined(__aarch64__) || defined(ARM_BF16_SUPPORT)
 template <>
 struct VecTypeTrait<c10::BFloat16> {
   using vec_t = vec_op::BF16Vec16;
 };
+#endif
 
 template <>
 struct VecTypeTrait<c10::Half> {
@@ -44,9 +50,21 @@ struct Counter {
 
 inline int64_t get_l2_size() {
   static int64_t size = []() {
+#if defined(__APPLE__)
+    // macOS doesn't have _SC_LEVEL2_CACHE_SIZE. Use sysctlbyname.
+    int64_t l2_cache_size = 0;
+    size_t len = sizeof(l2_cache_size);
+    if (sysctlbyname("hw.l2cachesize", &l2_cache_size, &len, NULL, 0) == 0 &&
+        l2_cache_size > 0) {
+      return l2_cache_size >> 1;  // use 50% of L2 cache
+    }
+    // Fallback if sysctlbyname fails
+    return 128LL * 1024 >> 1;  // use 50% of 128KB
+#else
     long l2_cache_size = sysconf(_SC_LEVEL2_CACHE_SIZE);
     assert(l2_cache_size != -1);
     return l2_cache_size >> 1;  // use 50% of L2 cache
+#endif
   }();
   return size;
 }