diff --git a/rust/.cargo/config.toml b/rust/.cargo/config.toml
index b327ff53f8..6102a36c00 100644
--- a/rust/.cargo/config.toml
+++ b/rust/.cargo/config.toml
@@ -30,11 +30,5 @@ rustflags = [
 [target.x86_64-unknown-linux-gnu]
 rustflags = ["-C", "target-cpu=haswell", "-C", "target-feature=+avx2,+fma,+f16c"]
 
-[target.x86_64-unknown-linux-gnu-avx512bf16]
-rustflags = ["-C", "target-feature=+avx2,+fma,+f16c,+avx512bf16"]
-
 [target.aarch64-apple-darwin]
 rustflags = ["-C", "target-cpu=apple-m1", "-C", "target-feature=+neon,+fp16,+fhm,+dotprod"]
-
-[target.aarch64-apple-darwin-m2]
-rustflags = ["-C", "target-cpu=apple-m2", "-C", "target-feature=+neon,+fp16,+fhm,+dotprod,+bf16"]
diff --git a/rust/lance-linalg/build.rs b/rust/lance-linalg/build.rs
index 3ba4ddeb68..29192df8ec 100644
--- a/rust/lance-linalg/build.rs
+++ b/rust/lance-linalg/build.rs
@@ -50,7 +50,8 @@ fn main() {
             .compile("f16");
     }
 
-    if cfg!(all(feature = "avx512bf16")) {
+    if cfg!(all(target_os = "linux", feature = "avx512bf16")) {
+        // No enable bf16 kernels on sappphire rapids
         cc::Build::new()
             .compiler("clang")
             .std("c17")
diff --git a/rust/lance-linalg/src/distance/l2.rs b/rust/lance-linalg/src/distance/l2.rs
index 5e1e03f413..a34a937364 100644
--- a/rust/lance-linalg/src/distance/l2.rs
+++ b/rust/lance-linalg/src/distance/l2.rs
@@ -118,7 +118,8 @@ impl L2 for BFloat16Type {
 
 #[cfg(any(
     all(target_os = "macos", target_feature = "neon"),
-    all(target_os = "linux", feature = "avx512fp16")
+    all(target_os = "linux", feature = "avx512fp16"),
+    all(target_os = "linux", feature = "avx512bf16")
 ))]
 mod kernel {
     use super::*;
diff --git a/rust/lance-linalg/src/distance/norm_l2.rs b/rust/lance-linalg/src/distance/norm_l2.rs
index 1483ab68b0..05f63a05ad 100644
--- a/rust/lance-linalg/src/distance/norm_l2.rs
+++ b/rust/lance-linalg/src/distance/norm_l2.rs
@@ -28,17 +28,19 @@ pub trait Normalize<T: Float> {
     fn norm_l2(&self) -> f32;
 }
 
-// `avx512fp16` is not supported in rustc yet. Once it is supported, we can
-// move it to target_feture.
+// `avx512fp16` and `avx512bf16` are not supported in rustc yet. Once they are
+// supported, we can move this to target_feture.
 #[cfg(any(
     all(target_os = "macos", target_feature = "neon"),
-    feature = "avx512fp16"
+    all(target_os = "linux", feature = "avx512fp16"),
+    all(target_os = "linux", feature = "avx512bf16"),
 ))]
 mod kernel {
     use super::*;
 
     extern "C" {
         pub fn norm_l2_f16(ptr: *const f16, len: u32) -> f32;
+        pub fn norm_l2_bf16(ptr: *const bf16, len: u32) -> f32;
     }
 }
 
@@ -88,32 +90,18 @@ impl Normalize<f16> for &[f16] {
     }
 }
 
-// `avx512bf16` is not supported in rustc yet. Once it is supported, we can
-// move it to target_feture.
-#[cfg(any(
-all(target_os = "macos", target_feature = "neon"),
-feature = "avx512bf16",
-))]
-mod kernel {
-    use super::*;
-
-    extern "C" {
-        pub fn norm_l2_bf16(ptr: *const bf16, len: u32) -> f32;
-    }
-}
-
 impl Normalize<bf16> for &[bf16] {
     fn norm_l2(&self) -> f32 {
         #[cfg(any(
             all(target_os = "macos", target_feature = "neon"),
-            feature = "avx512bf16",
+            all(target_os = "linux", target_feature = "avx512bf16"),
         ))]
         unsafe {
             kernel::norm_l2_bf16(self.as_ptr(), self.len() as u32)
         }
         #[cfg(not(any(
             all(target_os = "macos", target_feature = "neon"),
-            feature = "avx512bf16",
+            all(target_os = "linux", target_feature = "avx512bf16"),
         )))]
         norm_l2_impl::<bf16, 32>(self)
     }
diff --git a/rust/lance-linalg/src/simd/bf16.c b/rust/lance-linalg/src/simd/bf16.c
index 402500af01..4cfc926a94 100644
--- a/rust/lance-linalg/src/simd/bf16.c
+++ b/rust/lance-linalg/src/simd/bf16.c
@@ -16,10 +16,6 @@
 #include <stdint.h>
 #include <immintrin.h>
 
-// See https://github.com/ashvardanian/SimSIMD/blob/main/include/simsimd/spatial.h
-// https://www.intel.com/content/www/us/en/developer/articles/technical/intel-deep-learning-boost-new-instruction-bfloat16.html
-/// Needs avx512f, avx512bf16 and avx512vl
-
 /// @brief Computes the L2 norm of a bf16 vector.
 /// @param x A bf16 vector
 /// @param dimension The dimension of the vectors