diff --git a/include/cglm/affine-mat.h b/include/cglm/affine-mat.h
index 3fd19ef..e015591 100644
--- a/include/cglm/affine-mat.h
+++ b/include/cglm/affine-mat.h
@@ -81,6 +81,60 @@ glm_mul(mat4 m1, mat4 m2, mat4 dest) {
 #endif
 }
 
+/*!
+ * @brief this is similar to glm_mat4_mul but specialized to affine transform
+ *
+ * Right Matrix format should be:
+ *   R  R  R  0
+ *   R  R  R  0
+ *   R  R  R  0
+ *   0  0  0  1
+ *
+ * this reduces some multiplications. It should be faster than mat4_mul.
+ * if you are not sure about matrix format then DON'T use this! use mat4_mul
+ *
+ * @param[in]   m1    affine matrix 1
+ * @param[in]   m2    affine matrix 2
+ * @param[out]  dest  result matrix
+ */
+CGLM_INLINE
+void
+glm_mul_rot(mat4 m1, mat4 m2, mat4 dest) {
+#if defined( __SSE__ ) || defined( __SSE2__ )
+  glm_mul_rot_sse2(m1, m2, dest);
+#else
+  float a00 = m1[0][0], a01 = m1[0][1], a02 = m1[0][2], a03 = m1[0][3],
+        a10 = m1[1][0], a11 = m1[1][1], a12 = m1[1][2], a13 = m1[1][3],
+        a20 = m1[2][0], a21 = m1[2][1], a22 = m1[2][2], a23 = m1[2][3],
+        a30 = m1[3][0], a31 = m1[3][1], a32 = m1[3][2], a33 = m1[3][3],
+
+        b00 = m2[0][0], b01 = m2[0][1], b02 = m2[0][2],
+        b10 = m2[1][0], b11 = m2[1][1], b12 = m2[1][2],
+        b20 = m2[2][0], b21 = m2[2][1], b22 = m2[2][2],
+        b33 = m2[3][3];
+
+  dest[0][0] = a00 * b00 + a10 * b01 + a20 * b02;
+  dest[0][1] = a01 * b00 + a11 * b01 + a21 * b02;
+  dest[0][2] = a02 * b00 + a12 * b01 + a22 * b02;
+  dest[0][3] = a03 * b00 + a13 * b01 + a23 * b02;
+
+  dest[1][0] = a00 * b10 + a10 * b11 + a20 * b12;
+  dest[1][1] = a01 * b10 + a11 * b11 + a21 * b12;
+  dest[1][2] = a02 * b10 + a12 * b11 + a22 * b12;
+  dest[1][3] = a03 * b10 + a13 * b11 + a23 * b12;
+
+  dest[2][0] = a00 * b20 + a10 * b21 + a20 * b22;
+  dest[2][1] = a01 * b20 + a11 * b21 + a21 * b22;
+  dest[2][2] = a02 * b20 + a12 * b21 + a22 * b22;
+  dest[2][3] = a03 * b20 + a13 * b21 + a23 * b22;
+
+  dest[3][0] = a30;
+  dest[3][1] = a31;
+  dest[3][2] = a32;
+  dest[3][3] = a33;
+#endif
+}
+
 /*!
  * @brief inverse orthonormal rotation + translation matrix (ridig-body)
  *
diff --git a/include/cglm/simd/sse2/affine.h b/include/cglm/simd/sse2/affine.h
index 3ec8f5f..b746d0f 100644
--- a/include/cglm/simd/sse2/affine.h
+++ b/include/cglm/simd/sse2/affine.h
@@ -49,6 +49,38 @@ glm_mul_sse2(mat4 m1, mat4 m2, mat4 dest) {
                                      _mm_mul_ps(_mm_shuffle1_ps1(r, 3), l3))));
 }
 
+CGLM_INLINE
+void
+glm_mul_rot_sse2(mat4 m1, mat4 m2, mat4 dest) {
+  /* D = R * L (Column-Major) */
+  __m128 l0, l1, l2, l3, r;
+
+  l0 = _mm_load_ps(m1[0]);
+  l1 = _mm_load_ps(m1[1]);
+  l2 = _mm_load_ps(m1[2]);
+  l3 = _mm_load_ps(m1[3]);
+
+  r = _mm_load_ps(m2[0]);
+  _mm_store_ps(dest[0],
+               _mm_add_ps(_mm_add_ps(_mm_mul_ps(_mm_shuffle1_ps1(r, 0), l0),
+                                     _mm_mul_ps(_mm_shuffle1_ps1(r, 1), l1)),
+                          _mm_mul_ps(_mm_shuffle1_ps1(r, 2), l2)));
+
+  r = _mm_load_ps(m2[1]);
+  _mm_store_ps(dest[1],
+               _mm_add_ps(_mm_add_ps(_mm_mul_ps(_mm_shuffle1_ps1(r, 0), l0),
+                                     _mm_mul_ps(_mm_shuffle1_ps1(r, 1), l1)),
+                          _mm_mul_ps(_mm_shuffle1_ps1(r, 2), l2)));
+
+  r = _mm_load_ps(m2[2]);
+  _mm_store_ps(dest[2],
+               _mm_add_ps(_mm_add_ps(_mm_mul_ps(_mm_shuffle1_ps1(r, 0), l0),
+                                     _mm_mul_ps(_mm_shuffle1_ps1(r, 1), l1)),
+                          _mm_mul_ps(_mm_shuffle1_ps1(r, 2), l2)));
+
+  _mm_store_ps(dest[3], l3);
+}
+
 CGLM_INLINE
 void
 glm_inv_tr_sse2(mat4 mat) {