diff --git a/vp9/common/arm/neon/vp9_short_idct32x32_1_add_neon.asm b/vp9/common/arm/neon/vp9_short_idct32x32_1_add_neon.asm
new file mode 100644
index 0000000000000000000000000000000000000000..1f03ff60e005495fdb56d7eac66378770616ba1f
--- /dev/null
+++ b/vp9/common/arm/neon/vp9_short_idct32x32_1_add_neon.asm
@@ -0,0 +1,144 @@
+;
+; Copyright (c) 2013 The WebM project authors. All Rights Reserved.
+;
+; Use of this source code is governed by a BSD-style license and patent
+; grant that can be found in the LICENSE file in the root of the source
+; tree. All contributing project authors may be found in the AUTHORS
+; file in the root of the source tree.
+;
+
+ EXPORT |vp9_idct32x32_1_add_neon|
+ ARM
+ REQUIRE8
+ PRESERVE8
+
+ AREA ||.text||, CODE, READONLY, ALIGN=2
+
+ ;TODO(hkuang): put the following macros in a seperate
+ ;file so other idct function could also use them.
+ MACRO
+ LD_16x8 $src, $stride
+ vld1.8 {q8}, [$src], $stride
+ vld1.8 {q9}, [$src], $stride
+ vld1.8 {q10}, [$src], $stride
+ vld1.8 {q11}, [$src], $stride
+ vld1.8 {q12}, [$src], $stride
+ vld1.8 {q13}, [$src], $stride
+ vld1.8 {q14}, [$src], $stride
+ vld1.8 {q15}, [$src], $stride
+ MEND
+
+ MACRO
+ ADD_DIFF_16x8 $diff
+ vqadd.u8 q8, q8, $diff
+ vqadd.u8 q9, q9, $diff
+ vqadd.u8 q10, q10, $diff
+ vqadd.u8 q11, q11, $diff
+ vqadd.u8 q12, q12, $diff
+ vqadd.u8 q13, q13, $diff
+ vqadd.u8 q14, q14, $diff
+ vqadd.u8 q15, q15, $diff
+ MEND
+
+ MACRO
+ SUB_DIFF_16x8 $diff
+ vqsub.u8 q8, q8, $diff
+ vqsub.u8 q9, q9, $diff
+ vqsub.u8 q10, q10, $diff
+ vqsub.u8 q11, q11, $diff
+ vqsub.u8 q12, q12, $diff
+ vqsub.u8 q13, q13, $diff
+ vqsub.u8 q14, q14, $diff
+ vqsub.u8 q15, q15, $diff
+ MEND
+
+ MACRO
+ ST_16x8 $dst, $stride
+ vst1.8 {q8}, [$dst], $stride
+ vst1.8 {q9}, [$dst], $stride
+ vst1.8 {q10},[$dst], $stride
+ vst1.8 {q11},[$dst], $stride
+ vst1.8 {q12},[$dst], $stride
+ vst1.8 {q13},[$dst], $stride
+ vst1.8 {q14},[$dst], $stride
+ vst1.8 {q15},[$dst], $stride
+ MEND
+
+;void vp9_idct32x32_1_add_neon(int16_t *input, uint8_t *dest,
+; int dest_stride)
+;
+; r0 int16_t input
+; r1 uint8_t *dest
+; r2 int dest_stride
+
+|vp9_idct32x32_1_add_neon| PROC
+ push {lr}
+ pld [r1]
+ add r3, r1, #16 ; r3 dest + 16 for second loop
+ ldrsh r0, [r0]
+
+ ; generate cospi_16_64 = 11585
+ mov r12, #0x2d00
+ add r12, #0x41
+
+ ; out = dct_const_round_shift(input[0] * cospi_16_64)
+ mul r0, r0, r12 ; input[0] * cospi_16_64
+ add r0, r0, #0x2000 ; +(1 << ((DCT_CONST_BITS) - 1))
+ asr r0, r0, #14 ; >> DCT_CONST_BITS
+
+ ; out = dct_const_round_shift(out * cospi_16_64)
+ mul r0, r0, r12 ; out * cospi_16_64
+ mov r12, r1 ; save dest
+ add r0, r0, #0x2000 ; +(1 << ((DCT_CONST_BITS) - 1))
+ asr r0, r0, #14 ; >> DCT_CONST_BITS
+
+ ; a1 = ROUND_POWER_OF_TWO(out, 6)
+ add r0, r0, #32 ; + (1 <<((6) - 1))
+ asrs r0, r0, #6 ; >> 6
+ bge diff_positive_32_32
+
+diff_negative_32_32
+ neg r0, r0
+ usat r0, #8, r0
+ vdup.u8 q0, r0
+ mov r0, #4
+
+diff_negative_32_32_loop
+ sub r0, #1
+ LD_16x8 r1, r2
+ SUB_DIFF_16x8 q0
+ ST_16x8 r12, r2
+
+ LD_16x8 r1, r2
+ SUB_DIFF_16x8 q0
+ ST_16x8 r12, r2
+ cmp r0, #2
+ moveq r1, r3
+ moveq r12, r3
+ cmp r0, #0
+ bne diff_negative_32_32_loop
+ pop {r3,pc}
+
+diff_positive_32_32
+ usat r0, #8, r0
+ vdup.u8 q0, r0
+ mov r0, #4
+
+diff_positive_32_32_loop
+ sub r0, #1
+ LD_16x8 r1, r2
+ ADD_DIFF_16x8 q0
+ ST_16x8 r12, r2
+
+ LD_16x8 r1, r2
+ ADD_DIFF_16x8 q0
+ ST_16x8 r12, r2
+ cmp r0, #2
+ moveq r1, r3
+ moveq r12, r3
+ cmp r0, #0
+ bne diff_positive_32_32_loop
+ pop {pc}
+
+ ENDP ; |vp9_idct32x32_1_add_neon|
+ END
diff --git a/vp9/common/vp9_rtcd_defs.sh b/vp9/common/vp9_rtcd_defs.sh
index 5e049c63ce21e62e998d06b44e9a39cf68089b6b..ff73674b36cd4e5e0f86303888015e1f6201f3b5 100644
--- a/vp9/common/vp9_rtcd_defs.sh
+++ b/vp9/common/vp9_rtcd_defs.sh
@@ -299,7 +299,7 @@ prototype void vp9_idct32x32_34_add "const int16_t *input, uint8_t *dest, int de
specialize vp9_idct32x32_34_add sse2
prototype void vp9_idct32x32_1_add "const int16_t *input, uint8_t *dest, int dest_stride"
-specialize vp9_idct32x32_1_add sse2 dspr2
+specialize vp9_idct32x32_1_add sse2 neon dspr2
prototype void vp9_iht4x4_16_add "const int16_t *input, uint8_t *dest, int dest_stride, int tx_type"
specialize vp9_iht4x4_16_add sse2 neon dspr2
diff --git a/vp9/vp9_common.mk b/vp9/vp9_common.mk
index 11fa2e0c0e8a7479f17e78bb3bef6d1ccbc06a9b..c24ea023edadea077de565dba93416243a3e13e9 100644
--- a/vp9/vp9_common.mk
+++ b/vp9/vp9_common.mk
@@ -122,6 +122,7 @@ VP9_COMMON_SRCS-$(HAVE_NEON) += common/arm/neon/vp9_short_idct8x8_1_add_neon$(AS
VP9_COMMON_SRCS-$(HAVE_NEON) += common/arm/neon/vp9_short_idct8x8_add_neon$(ASM)
VP9_COMMON_SRCS-$(HAVE_NEON) += common/arm/neon/vp9_short_idct16x16_1_add_neon$(ASM)
VP9_COMMON_SRCS-$(HAVE_NEON) += common/arm/neon/vp9_short_idct16x16_add_neon$(ASM)
+VP9_COMMON_SRCS-$(HAVE_NEON) += common/arm/neon/vp9_short_idct32x32_1_add_neon$(ASM)
VP9_COMMON_SRCS-$(HAVE_NEON) += common/arm/neon/vp9_short_idct32x32_add_neon$(ASM)
VP9_COMMON_SRCS-$(HAVE_NEON) += common/arm/neon/vp9_short_iht4x4_add_neon$(ASM)
VP9_COMMON_SRCS-$(HAVE_NEON) += common/arm/neon/vp9_short_iht8x8_add_neon$(ASM)