diff --git a/tz/avx2_amd64.s b/tz/avx2_amd64.s
index 9aa8a70..787dfc7 100644
--- a/tz/avx2_amd64.s
+++ b/tz/avx2_amd64.s
@@ -18,14 +18,14 @@
     VPXOR Y3, Y8, Y3 \
     VPAND Y3, Y2, Y4 \
     VPXOR Y4, Y0, Y8 \
-    VMOVDQA Y3, Y0
+    VMOVDQU Y3, Y0
 
 // func mulByteRightx2(c00c10, c01c11 *[4]uint64, b byte)
 TEXT ·mulByteRightx2(SB),NOSPLIT,$0
     MOVQ c00c10+0(FP), AX
-    VMOVDQA (AX), Y0
+    VMOVDQU (AX), Y0
     MOVQ c01c11+8(FP), BX
-    VMOVDQA (BX), Y8
+    VMOVDQU (BX), Y8
 
     VPXOR Y13, Y13, Y13    // Y13 = 0x0000...
     VPCMPEQB Y12, Y12, Y12 // Y12 = 0xFFFF...
@@ -43,17 +43,17 @@ TEXT ·mulByteRightx2(SB),NOSPLIT,$0
     mulBit($1)
     mulBit($0)
 
-    VMOVDQA Y8, (BX)
-    VMOVDQA Y0, (AX)
+    VMOVDQU Y8, (BX)
+    VMOVDQU Y0, (AX)
 
     RET
 
 // func mulBitRightx2(c00c10, c01c11 *[4]uint64, e *[2]uint64)
 TEXT ·mulBitRightx2(SB),NOSPLIT,$0
     MOVQ c00c10+0(FP), AX
-    VMOVDQA (AX), Y0
+    VMOVDQU (AX), Y0
     MOVQ c01c11+8(FP), BX
-    VMOVDQA (BX), Y8
+    VMOVDQU (BX), Y8
 
     VPSLLQ $1, Y0, Y1
     VPALIGNR $8, Y1, Y0, Y2
@@ -70,6 +70,6 @@ TEXT ·mulBitRightx2(SB),NOSPLIT,$0
     VPXOR Y3, Y8, Y3
     VPAND Y3, Y2, Y4
     VPXOR Y4, Y0, Y8
-    VMOVDQA Y8, (BX)
-    VMOVDQA Y3, (AX)
+    VMOVDQU Y8, (BX)
+    VMOVDQU Y3, (AX)
     RET