diff --git a/aom_dsp/x86/blend_a64_mask_avx2.c b/aom_dsp/x86/blend_a64_mask_avx2.c
index 09c63a3..dfbab32 100644
--- a/aom_dsp/x86/blend_a64_mask_avx2.c
+++ b/aom_dsp/x86/blend_a64_mask_avx2.c
@@ -1307,7 +1307,7 @@
   const __m256i v_round_offset = _mm256_set1_epi32(round_offset);
   const int shift = round_bits + AOM_BLEND_A64_ROUND_BITS;
 
-  const __m256i clip_low = _mm256_set1_epi16(0);
+  const __m256i clip_low = _mm256_setzero_si256();
   const __m256i clip_high = _mm256_set1_epi16((1 << bd) - 1);
   const __m256i mask_max = _mm256_set1_epi16(AOM_BLEND_A64_MAX_ALPHA);
 
diff --git a/aom_dsp/x86/blend_a64_mask_sse4.c b/aom_dsp/x86/blend_a64_mask_sse4.c
index 270ad26..58a7345 100644
--- a/aom_dsp/x86/blend_a64_mask_sse4.c
+++ b/aom_dsp/x86/blend_a64_mask_sse4.c
@@ -1493,7 +1493,7 @@
   const __m128i v_round_offset = _mm_set1_epi32(round_offset);
   const int shift = round_bits + AOM_BLEND_A64_ROUND_BITS;
 
-  const __m128i clip_low = _mm_set1_epi16(0);
+  const __m128i clip_low = _mm_setzero_si128();
   const __m128i clip_high = _mm_set1_epi16((1 << bd) - 1);
   const __m128i mask_max = _mm_set1_epi16(AOM_BLEND_A64_MAX_ALPHA);
 
diff --git a/aom_dsp/x86/loopfilter_avx2.c b/aom_dsp/x86/loopfilter_avx2.c
index b593819..af6c5da 100644
--- a/aom_dsp/x86/loopfilter_avx2.c
+++ b/aom_dsp/x86/loopfilter_avx2.c
@@ -32,7 +32,7 @@
       _mm_broadcastb_epi8(_mm_cvtsi32_si128((int)_limit0[0]));
   const __m128i blimit_v =
       _mm_broadcastb_epi8(_mm_cvtsi32_si128((int)_blimit0[0]));
-  const __m128i zero = _mm_set1_epi16(0);
+  const __m128i zero = _mm_setzero_si128();
   const __m128i ff = _mm_cmpeq_epi8(zero, zero);
 
   p256_2 =
@@ -239,7 +239,7 @@
       _mm_broadcastb_epi8(_mm_cvtsi32_si128((int)_limit0[0]));
   const __m128i blimit_v =
       _mm_broadcastb_epi8(_mm_cvtsi32_si128((int)_blimit0[0]));
-  const __m128i zero = _mm_set1_epi16(0);
+  const __m128i zero = _mm_setzero_si128();
   const __m128i ff = _mm_cmpeq_epi8(zero, zero);
 
   p256_3 =
@@ -486,7 +486,7 @@
       _mm_broadcastb_epi8(_mm_cvtsi32_si128((int)_limit0[0]));
   const __m128i blimit_v =
       _mm_broadcastb_epi8(_mm_cvtsi32_si128((int)_blimit0[0]));
-  const __m128i zero = _mm_set1_epi16(0);
+  const __m128i zero = _mm_setzero_si128();
   const __m128i ff = _mm_cmpeq_epi8(zero, zero);
 
   p256_3 =
diff --git a/aom_dsp/x86/loopfilter_sse2.c b/aom_dsp/x86/loopfilter_sse2.c
index 87c5bb3..731dd10 100644
--- a/aom_dsp/x86/loopfilter_sse2.c
+++ b/aom_dsp/x86/loopfilter_sse2.c
@@ -2133,7 +2133,7 @@
                                      const unsigned char *_blimit0,
                                      const unsigned char *_limit0,
                                      const unsigned char *_thresh0) {
-  const __m128i zero = _mm_set1_epi16(0);
+  const __m128i zero = _mm_setzero_si128();
   const __m128i one = _mm_set1_epi8(1);
   const __m128i blimit_v = _mm_load_si128((const __m128i *)_blimit0);
   const __m128i limit_v = _mm_load_si128((const __m128i *)_limit0);
@@ -2438,7 +2438,7 @@
                                     const unsigned char *_blimit0,
                                     const unsigned char *_limit0,
                                     const unsigned char *_thresh0) {
-  const __m128i zero = _mm_set1_epi16(0);
+  const __m128i zero = _mm_setzero_si128();
   const __m128i one = _mm_set1_epi8(1);
   const __m128i blimit_v = _mm_load_si128((const __m128i *)_blimit0);
   const __m128i limit_v = _mm_load_si128((const __m128i *)_limit0);
@@ -2630,7 +2630,7 @@
                                     const unsigned char *_blimit0,
                                     const unsigned char *_limit0,
                                     const unsigned char *_thresh0) {
-  const __m128i zero = _mm_set1_epi16(0);
+  const __m128i zero = _mm_setzero_si128();
   const __m128i one = _mm_set1_epi8(1);
   const __m128i blimit_v = _mm_load_si128((const __m128i *)_blimit0);
   const __m128i limit_v = _mm_load_si128((const __m128i *)_limit0);
@@ -2802,7 +2802,7 @@
                                     const unsigned char *_blimit0,
                                     const unsigned char *_limit0,
                                     const unsigned char *_thresh0) {
-  const __m128i zero = _mm_set1_epi16(0);
+  const __m128i zero = _mm_setzero_si128();
   const __m128i blimit_v = _mm_load_si128((const __m128i *)_blimit0);
   const __m128i limit_v = _mm_load_si128((const __m128i *)_limit0);
   const __m128i thresh_v = _mm_load_si128((const __m128i *)_thresh0);
diff --git a/aom_dsp/x86/quantize_avx2.c b/aom_dsp/x86/quantize_avx2.c
index 5763bd6..b808d46 100644
--- a/aom_dsp/x86/quantize_avx2.c
+++ b/aom_dsp/x86/quantize_avx2.c
@@ -128,7 +128,7 @@
                          const int16_t *iscan) {
   (void)scan;
   __m256i v_zbin, v_round, v_quant, v_dequant, v_quant_shift;
-  __m256i v_eobmax = _mm256_set1_epi16(0);
+  __m256i v_eobmax = _mm256_setzero_si256();
 
   load_b_values_avx2(zbin_ptr, &v_zbin, round_ptr, &v_round, quant_ptr,
                      &v_quant, dequant_ptr, &v_dequant, quant_shift_ptr,
@@ -211,7 +211,7 @@
     tran_low_t *dqcoeff_ptr, const int16_t *dequant_ptr, uint16_t *eob_ptr,
     const int16_t *iscan, int log_scale) {
   __m256i v_zbin, v_round, v_quant, v_dequant, v_quant_shift;
-  __m256i v_eobmax = _mm256_set1_epi16(0);
+  __m256i v_eobmax = _mm256_setzero_si256();
 
   load_b_values_avx2(zbin_ptr, &v_zbin, round_ptr, &v_round, quant_ptr,
                      &v_quant, dequant_ptr, &v_dequant, quant_shift_ptr,
diff --git a/aom_dsp/x86/variance_impl_avx2.c b/aom_dsp/x86/variance_impl_avx2.c
index 8f11ada..8ea0443 100644
--- a/aom_dsp/x86/variance_impl_avx2.c
+++ b/aom_dsp/x86/variance_impl_avx2.c
@@ -171,9 +171,9 @@
   __m256i sse_reg, sum_reg, sse_reg_hi, res_cmp, sum_reg_lo, sum_reg_hi;
   __m256i zero_reg;
   int i, sum;
-  sum_reg = _mm256_set1_epi16(0);
-  sse_reg = _mm256_set1_epi16(0);
-  zero_reg = _mm256_set1_epi16(0);
+  sum_reg = _mm256_setzero_si256();
+  sse_reg = _mm256_setzero_si256();
+  zero_reg = _mm256_setzero_si256();
 
   // x_offset = 0 and y_offset = 0
   if (x_offset == 0) {
@@ -359,9 +359,9 @@
   __m256i sse_reg, sum_reg, sse_reg_hi, res_cmp, sum_reg_lo, sum_reg_hi;
   __m256i zero_reg;
   int i, sum;
-  sum_reg = _mm256_set1_epi16(0);
-  sse_reg = _mm256_set1_epi16(0);
-  zero_reg = _mm256_set1_epi16(0);
+  sum_reg = _mm256_setzero_si256();
+  sse_reg = _mm256_setzero_si256();
+  zero_reg = _mm256_setzero_si256();
 
   // x_offset = 0 and y_offset = 0
   if (x_offset == 0) {
@@ -599,9 +599,9 @@
   __m256i sse_reg, sum_reg, sse_reg_hi, res_cmp, sum_reg_lo, sum_reg_hi;
   __m256i zero_reg;
   int i, sum;
-  sum_reg = _mm256_set1_epi16(0);
-  sse_reg = _mm256_set1_epi16(0);
-  zero_reg = _mm256_set1_epi16(0);
+  sum_reg = _mm256_setzero_si256();
+  sse_reg = _mm256_setzero_si256();
+  zero_reg = _mm256_setzero_si256();
 
   // x_offset = 0 and y_offset = 0
   if (x_offset == 0) {
diff --git a/av1/common/x86/highbd_convolve_2d_avx2.c b/av1/common/x86/highbd_convolve_2d_avx2.c
index 12046e4..429f8f1 100644
--- a/av1/common/x86/highbd_convolve_2d_avx2.c
+++ b/av1/common/x86/highbd_convolve_2d_avx2.c
@@ -80,7 +80,7 @@
       for (i = 0; i < im_h; i += 2) {
         const __m256i row0 =
             _mm256_loadu_si256((__m256i *)&src_ptr[i * src_stride + j]);
-        __m256i row1 = _mm256_set1_epi16(0);
+        __m256i row1 = _mm256_setzero_si256();
         if (i + 1 < im_h)
           row1 =
               _mm256_loadu_si256((__m256i *)&src_ptr[(i + 1) * src_stride + j]);
diff --git a/av1/common/x86/highbd_inv_txfm_sse4.c b/av1/common/x86/highbd_inv_txfm_sse4.c
index 37f8f42..de3af3a 100644
--- a/av1/common/x86/highbd_inv_txfm_sse4.c
+++ b/av1/common/x86/highbd_inv_txfm_sse4.c
@@ -517,7 +517,7 @@
 static void iadst4x4_sse4_1(__m128i *in, __m128i *out, int bit, int do_cols,
                             int bd, int out_shift) {
   const int32_t *sinpi = sinpi_arr(bit);
-  const __m128i zero = _mm_set1_epi32(0);
+  const __m128i zero = _mm_setzero_si128();
   __m128i rnding = _mm_set1_epi32(1 << (bit + 4 - 1));
   rnding = _mm_unpacklo_epi32(rnding, zero);
   const __m128i mul = _mm_set1_epi32(1 << 4);
@@ -698,7 +698,7 @@
                               int bd, int out_shift) {
   (void)bit;
   __m128i v[4];
-  __m128i zero = _mm_set1_epi32(0);
+  __m128i zero = _mm_setzero_si128();
   __m128i fact = _mm_set1_epi32(NewSqrt2);
   __m128i offset = _mm_set1_epi32(1 << (NewSqrt2Bits - 1));
   __m128i a0_low, a1_low;
@@ -3142,7 +3142,7 @@
   __m128i fact = _mm_set1_epi32(2 * NewSqrt2);
   __m128i offset = _mm_set1_epi32(1 << (NewSqrt2Bits - 1));
   __m128i a0_low, a0_high, a1_low, a1_high;
-  __m128i zero = _mm_set1_epi32(0);
+  __m128i zero = _mm_setzero_si128();
   offset = _mm_unpacklo_epi32(offset, zero);
 
   for (int i = 0; i < 16; i++) {
diff --git a/av1/common/x86/highbd_jnt_convolve_avx2.c b/av1/common/x86/highbd_jnt_convolve_avx2.c
index 9cedd44..da52ecd 100644
--- a/av1/common/x86/highbd_jnt_convolve_avx2.c
+++ b/av1/common/x86/highbd_jnt_convolve_avx2.c
@@ -282,7 +282,7 @@
       for (i = 0; i < im_h; i += 2) {
         const __m256i row0 =
             _mm256_loadu_si256((__m256i *)&src_ptr[i * src_stride + j]);
-        __m256i row1 = _mm256_set1_epi16(0);
+        __m256i row1 = _mm256_setzero_si256();
         if (i + 1 < im_h)
           row1 =
               _mm256_loadu_si256((__m256i *)&src_ptr[(i + 1) * src_stride + j]);
diff --git a/av1/common/x86/highbd_warp_affine_avx2.c b/av1/common/x86/highbd_warp_affine_avx2.c
index 87b1a66..7f6aceb 100644
--- a/av1/common/x86/highbd_warp_affine_avx2.c
+++ b/av1/common/x86/highbd_warp_affine_avx2.c
@@ -158,7 +158,7 @@
             iy = iy * stride;
 
             __m256i v_refl = _mm256_inserti128_si256(
-                _mm256_set1_epi16(0),
+                _mm256_setzero_si256(),
                 _mm_loadu_si128((__m128i *)&ref[iy + ix4 - 7]), 0);
             v_refl = _mm256_inserti128_si256(
                 v_refl, _mm_loadu_si128((__m128i *)&ref[iy + ix4 + 1]),
@@ -218,7 +218,7 @@
                 _mm_shuffle_epi32(v_01, 3));  // A7A6A7A6A7A6A7A6
 
             __m256i v_refl = _mm256_inserti128_si256(
-                _mm256_set1_epi16(0),
+                _mm256_setzero_si256(),
                 _mm_loadu_si128((__m128i *)&ref[iy + ix4 - 7]), 0);
             v_refl = _mm256_inserti128_si256(
                 v_refl, _mm_loadu_si128((__m128i *)&ref[iy + ix4 + 1]),
@@ -333,7 +333,7 @@
             iy = iy * stride;
 
             __m256i v_refl = _mm256_inserti128_si256(
-                _mm256_set1_epi16(0),
+                _mm256_setzero_si256(),
                 _mm_loadu_si128((__m128i *)&ref[iy + ix4 - 7]), 0);
             v_refl = _mm256_inserti128_si256(
                 v_refl, _mm_loadu_si128((__m128i *)&ref[iy + ix4 + 1]),
@@ -454,7 +454,7 @@
                 _mm256_unpackhi_epi64(v_c0123u, v_c4567u);  // H7H6 ... A7A6
 
             __m256i v_refl = _mm256_inserti128_si256(
-                _mm256_set1_epi16(0),
+                _mm256_setzero_si256(),
                 _mm_loadu_si128((__m128i *)&ref[iy + ix4 - 7]), 0);
             v_refl = _mm256_inserti128_si256(
                 v_refl, _mm_loadu_si128((__m128i *)&ref[iy + ix4 + 1]),
diff --git a/av1/common/x86/warp_plane_avx2.c b/av1/common/x86/warp_plane_avx2.c
index b0c9a93..ceb836e 100644
--- a/av1/common/x86/warp_plane_avx2.c
+++ b/av1/common/x86/warp_plane_avx2.c
@@ -1028,12 +1028,12 @@
   int64_t sum_error = 0;
   int i, j;
   __m256i row_error, col_error;
-  __m256i zero = _mm256_set1_epi16(0);
+  __m256i zero = _mm256_setzero_si256();
   __m256i dup_255 = _mm256_set1_epi16(255);
   col_error = zero;
 
   for (i = 0; i < (p_height / 4); i++) {
-    row_error = _mm256_set1_epi16(0);
+    row_error = _mm256_setzero_si256();
     for (j = 0; j < (p_width / 16); j++) {
       __m256i ref_1_16 = _mm256_cvtepu8_epi16(_mm_load_si128(
           (__m128i *)(ref + (j * 16) + (((i * 4) + 0) * ref_stride))));
diff --git a/av1/common/x86/warp_plane_sse2.c b/av1/common/x86/warp_plane_sse2.c
index 6ff6665..f8fe578 100644
--- a/av1/common/x86/warp_plane_sse2.c
+++ b/av1/common/x86/warp_plane_sse2.c
@@ -21,7 +21,7 @@
   int64_t sum_error = 0;
   int i, j;
   __m128i row_error, col_error;
-  __m128i zero = _mm_set1_epi16(0);
+  __m128i zero = _mm_setzero_si128();
   __m128i dup_255 = _mm_set1_epi16(255);
   col_error = zero;
   for (i = 0; i < (p_height); i++) {
diff --git a/av1/encoder/x86/av1_fwd_txfm_sse2.c b/av1/encoder/x86/av1_fwd_txfm_sse2.c
index 694e613..a5cbe16 100644
--- a/av1/encoder/x86/av1_fwd_txfm_sse2.c
+++ b/av1/encoder/x86/av1_fwd_txfm_sse2.c
@@ -1425,7 +1425,7 @@
   const __m128i sinpi_p03_p04 = pair_set_epi16(sinpi[3], sinpi[4]);
   const __m128i sinpi_m03_p02 = pair_set_epi16(-sinpi[3], sinpi[2]);
   const __m128i sinpi_p03_p03 = _mm_set1_epi16((int16_t)sinpi[3]);
-  const __m128i __zero = _mm_set1_epi16(0);
+  const __m128i __zero = _mm_setzero_si128();
   const __m128i __rounding = _mm_set1_epi32(1 << (cos_bit - 1));
   const __m128i in7 = _mm_add_epi16(input[0], input[1]);
   __m128i u[8], v[8];
@@ -1573,7 +1573,7 @@
   const __m128i sinpi_p03_p04 = pair_set_epi16(sinpi[3], sinpi[4]);
   const __m128i sinpi_m03_p02 = pair_set_epi16(-sinpi[3], sinpi[2]);
   const __m128i sinpi_p03_p03 = _mm_set1_epi16((int16_t)sinpi[3]);
-  const __m128i __zero = _mm_set1_epi16(0);
+  const __m128i __zero = _mm_setzero_si128();
   const __m128i __rounding = _mm_set1_epi32(1 << (cos_bit - 1));
   const __m128i in7 = _mm_add_epi16(input[0], input[1]);
   __m128i u_lo[8], u_hi[8], v_lo[8], v_hi[8];