diff --git a/CMakeLists.txt b/CMakeLists.txt
index da50691..c8f847f 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -21,5 +21,8 @@ list(FILTER SOURCES EXCLUDE REGEX ".*src/main\\.c$")
 foreach(test_src ${TEST_SOURCES})
     get_filename_component(test_name ${test_src} NAME_WE)
     add_executable(${test_name} ${test_src} ${SOURCES})
+    if(NOT MSVC)
+        target_link_libraries(${test_name} m)
+    endif()
     add_test(NAME ${test_name} COMMAND ${test_name})
 endforeach()
\ No newline at end of file
diff --git a/src/main.c b/src/main.c
index 09d4445..f277e94 100644
--- a/src/main.c
+++ b/src/main.c
@@ -1,7 +1,13 @@
 #include <stdio.h>
-#include "math/vec4.h"
+#include "math/mat4.h"
 
 int main(void) 
 {
+    float arr[16] = {1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16};
+    Mat4f_t mat1 =  mat4f_from_array(arr);
+    Mat4f_t mat2 =  mat4f_scalar(1);
+    Mat4f_t mat3 =  mat4f_zero();
+    Mat4f_t mat4 =  mat4f_identity();
+    printf("%f", arr[1]);
     return 0;
 }
diff --git a/src/math/mat4.c b/src/math/mat4.c
index 70c92e2..e49fd20 100644
--- a/src/math/mat4.c
+++ b/src/math/mat4.c
@@ -1,27 +1,69 @@
-// #include "mat4.h"
-// #include "math.h"
+#include "mat4.h"
 
-// Mat4_t mat4(const float arr[16]) 
-// {
-//     Mat4_t mat;
-//     memcpy(mat.m, arr, 16*sizeof(float));
-//     return mat;
-// }
+Mat4f_t mat4f_from_array(const float arr[16])
+{
+    Mat4f_t mat;
+    for(int i = 0; i<MAT_SIZE; i+=4) {
+#if defined (SIMD_X86)
+        __m128 line = _mm_load_ps(&arr[i]);
+        _mm_store_ps(&mat.m[i], line);
+#elif defined (SIMD_ARCH)
+        float32x4_t line = vld1q_f32(&arr[i]);
+        vst1q_f32(&mat.m[i], line);
+#else
+        for(int j = 0; j<4; j++) {
+            mat.m[i+j] = arr[i+j];
+        }
+#endif
+    }
+    return mat;
+}
 
-// Mat4_t mat4_zro(void) 
-// {
-//     return (Mat4_t){0};
-// }
+Mat4f_t mat4f_scalar(float f)
+{
+    Mat4f_t mat;
+    for(int i = 0; i<MAT_SIZE; i+=4) {
+        #if defined (SIMD_X86)
+        __m128 line_scalar = _mm_set1_ps(f);
+        _mm_store_ps(&mat.m[i], line_scalar);
+        #elif defined (SIMD_ARCH)
+        float32x4_t line_scalar = vdupq_n_f32(f);
+        vst1q_f32(&mat.m[i], line_scalar);
+        #else
+        for(int j = 0; j<4; j++) {
+            mat.m[i+j] = f;
+        }
+        #endif
+    }
+    return mat;
+}
+
+Mat4f_t mat4f_zero()
+{
+#if defined (SIMD_X86)
+    Mat4f_t mat;
+    
+    for(int i = 0; i<MAT_SIZE; i+=4) {
+        __m128 line_zero = _mm_setzero_ps();
+        _mm_store_ps(&mat.m[i], line_zero);
+    }
+
+    return mat;
+#else
+    return mat4f_scalar(0.f);
+#endif
+}
+
+Mat4f_t mat4f_identity()
+{
+    Mat4f_t mat = mat4f_zero();
+    mat.m[0] = 1.f;
+    mat.m[5] = 1.f;
+    mat.m[10] = 1.f;
+    mat.m[15] = 1.f;
+    return mat;
+}
 
-// Mat4_t mat4_ity(void) 
-// {
-//     return (Mat4_t) {{
-//         1, 0, 0, 0, 
-//         0, 1, 0, 0,
-//         0, 0, 1, 0,
-//         0, 0, 0, 1,
-//     }};
-// }
 
 // Mat4_t mat4_add(const Mat4_t* m1, const Mat4_t* m2) 
 // {
diff --git a/src/math/mat4.h b/src/math/mat4.h
index de0f96b..65fc108 100644
--- a/src/math/mat4.h
+++ b/src/math/mat4.h
@@ -1,30 +1,37 @@
 #ifndef MATRIX4_H
 #define MATRIX4_H
 
+#include "mconfig.h"
+
+#define MAT_SIZE 16
+
 typedef struct 
 {
-    float m[16];
-} Mat4_t;
+    float m[MAT_SIZE];
+} ALIGN16 Mat4f_t;
 
-Mat4_t mat4(const float arr[16]);
+Mat4f_t mat4f_from_array(const float arr[16]);
+Mat4f_t mat4f_scalar(float f);
+Mat4f_t mat4f_zero(void);
+Mat4f_t mat4f_identity(void);
 
-Mat4_t mat4_zro(void);
+// Mat4_t mat4_zro(void);
 
-Mat4_t mat4_ity(void);
+// Mat4_t mat4_ity(void);
 
-Mat4_t mat4_add(const Mat4_t* m1, const Mat4_t* m2);
+// Mat4_t mat4_add(const Mat4_t* m1, const Mat4_t* m2);
 
-Mat4_t mat4_sub(const Mat4_t* m1, const Mat4_t* m2);
+// Mat4_t mat4_sub(const Mat4_t* m1, const Mat4_t* m2);
 
-Mat4_t mat4_scl(const Mat4_t* m, float scalar);
+// Mat4_t mat4_scl(const Mat4_t* m, float scalar);
 
-// row * col
-Mat4_t mat4_mul(const Mat4_t* m1, const Mat4_t* m2);
+// // row * col
+// Mat4_t mat4_mul(const Mat4_t* m1, const Mat4_t* m2);
 
-Mat4_t mat4_tpo(const Mat4_t* m);
+// Mat4_t mat4_tpo(const Mat4_t* m);
 
-float mat4_det(const Mat4_t* m);
+// float mat4_det(const Mat4_t* m);
 
-Mat4_t mat4_inv(const Mat4_t* m);
+// Mat4_t mat4_inv(const Mat4_t* m);
 
 #endif // MATRIX4_H
diff --git a/src/math/vec4.c b/src/math/vec4.c
index b8771a6..bb1547a 100644
--- a/src/math/vec4.c
+++ b/src/math/vec4.c
@@ -1,7 +1,5 @@
 #include "vec4.h"
 
-#define VEC_SIZE 4
-
 Vec4f_t vec4f(float x, float y, float z, float w) 
 {
     return (Vec4f_t){.x = x, .y = y, .z = z, .w = w};
diff --git a/src/math/vec4.h b/src/math/vec4.h
index c6ccc32..3e415ca 100644
--- a/src/math/vec4.h
+++ b/src/math/vec4.h
@@ -3,11 +3,13 @@
 
 #include "mconfig.h"
 
+#define VEC_SIZE 4
+
 // must be aligned by 16 Bytes (less instruction executed for SSE)
 typedef union
 {
     struct { float x, y, z, w; };
-    float data[4];
+    float data[VEC_SIZE];
 } ALIGN16 Vec4f_t;
 
 Vec4f_t vec4f_from_array(const float *__restrict val);