diff --git a/src/opts/SkColor_opts_neon.h b/src/opts/SkColor_opts_neon.h
index 85752f5558..a3430b5c51 100644
--- a/src/opts/SkColor_opts_neon.h
+++ b/src/opts/SkColor_opts_neon.h
@@ -82,4 +82,33 @@ static inline uint8x8_t SkBlend32_neon8(uint8x8_t src, uint8x8_t dst, uint16x8_t
     return vmovn_u16(vreinterpretq_u16_s16(dst_wide));
 }
 
+static inline SkPMColor SkFourByteInterp256_neon(SkPMColor src, SkPMColor dst,
+                                                 unsigned srcScale) {
+    SkASSERT(srcScale <= 256);
+    int16x8_t vscale = vdupq_n_s16(srcScale);
+    int16x8_t vsrc_wide, vdst_wide, vdiff;
+    uint8x8_t res;
+
+    vsrc_wide = vreinterpretq_s16_u16(vmovl_u8(vreinterpret_u8_u32(vdup_n_u32(src))));
+    vdst_wide = vreinterpretq_s16_u16(vmovl_u8(vreinterpret_u8_u32(vdup_n_u32(dst))));
+
+    vdiff = vsrc_wide - vdst_wide;
+    vdiff *= vscale;
+
+    vdiff = vshrq_n_s16(vdiff, 8);
+
+    vdst_wide += vdiff;
+
+    res = vmovn_u16(vreinterpretq_u16_s16(vdst_wide));
+
+    return vget_lane_u32(vreinterpret_u32_u8(res), 0);
+}
+
+static inline SkPMColor SkFourByteInterp_neon(SkPMColor src, SkPMColor dst,
+                                              U8CPU srcWeight) {
+    SkASSERT(srcWeight <= 255);
+    unsigned scale = SkAlpha255To256(srcWeight);
+    return SkFourByteInterp256_neon(src, dst, scale);
+}
+
 #endif /* #ifndef SkColor_opts_neon_DEFINED */
diff --git a/src/opts/SkXfermode_opts_arm_neon.cpp b/src/opts/SkXfermode_opts_arm_neon.cpp
index b8d8ef521b..7435dd44de 100644
--- a/src/opts/SkXfermode_opts_arm_neon.cpp
+++ b/src/opts/SkXfermode_opts_arm_neon.cpp
@@ -632,7 +632,7 @@ void SkNEONProcCoeffXfermode::xfer32(SkPMColor dst[], const SkPMColor src[],
                 SkPMColor dstC = dst[i];
                 SkPMColor C = proc(src[i], dstC);
                 if (a != 0xFF) {
-                    C = SkFourByteInterp(C, dstC, a);
+                    C = SkFourByteInterp_neon(C, dstC, a);
                 }
                 dst[i] = C;
             }
@@ -700,7 +700,7 @@ void SkNEONProcCoeffXfermode::xfer16(uint16_t* SK_RESTRICT dst,
                 SkPMColor dstC = SkPixel16ToPixel32(dst[i]);
                 SkPMColor C = proc(src[i], dstC);
                 if (0xFF != a) {
-                    C = SkFourByteInterp(C, dstC, a);
+                    C = SkFourByteInterp_neon(C, dstC, a);
                 }
                 dst[i] = SkPixel32ToPixel16_ToU16(C);
             }