diff --git a/vp8/common/rtcd_defs.pl b/vp8/common/rtcd_defs.pl
index d65ca604d..7924ae750 100644
--- a/vp8/common/rtcd_defs.pl
+++ b/vp8/common/rtcd_defs.pl
@@ -254,19 +254,6 @@ if ($opts{arch} =~ /x86/) {
     specialize qw/vp8_copy32xn sse2 sse3/;
 }
 
-#
-# Structured Similarity (SSIM)
-#
-if (vpx_config("CONFIG_INTERNAL_STATS") eq "yes") {
-    $opts{arch} eq "x86_64" and $sse2_on_x86_64 = "sse2";
-
-    add_proto qw/void vp8_ssim_parms_8x8/, "unsigned char *s, int sp, unsigned char *r, int rp, unsigned long *sum_s, unsigned long *sum_r, unsigned long *sum_sq_s, unsigned long *sum_sq_r, unsigned long *sum_sxr";
-    specialize qw/vp8_ssim_parms_8x8/, "$sse2_on_x86_64";
-
-    add_proto qw/void vp8_ssim_parms_16x16/, "unsigned char *s, int sp, unsigned char *r, int rp, unsigned long *sum_s, unsigned long *sum_r, unsigned long *sum_sq_s, unsigned long *sum_sq_r, unsigned long *sum_sxr";
-    specialize qw/vp8_ssim_parms_16x16/, "$sse2_on_x86_64";
-}
-
 #
 # Forward DCT
 #
diff --git a/vp8/encoder/onyx_if.c b/vp8/encoder/onyx_if.c
index c2a7ac4ce..d2fb05ab7 100644
--- a/vp8/encoder/onyx_if.c
+++ b/vp8/encoder/onyx_if.c
@@ -74,26 +74,7 @@ extern const int vp8_gf_interval_table[101];
 
 #if CONFIG_INTERNAL_STATS
 #include "math.h"
-
-extern double vp8_calc_ssim
-(
-    YV12_BUFFER_CONFIG *source,
-    YV12_BUFFER_CONFIG *dest,
-    int lumamask,
-    double *weight
-);
-
-
-extern double vp8_calc_ssimg
-(
-    YV12_BUFFER_CONFIG *source,
-    YV12_BUFFER_CONFIG *dest,
-    double *ssim_y,
-    double *ssim_u,
-    double *ssim_v
-);
-
-
+#include "vpx_dsp/ssim.h"
 #endif
 
 
@@ -5741,8 +5722,8 @@ int vp8_get_compressed_data(VP8_COMP *cpi, unsigned int *frame_flags, unsigned l
                     cpi->total_sq_error2 += sq_error2;
                     cpi->totalp  += frame_psnr2;
 
-                    frame_ssim2 = vp8_calc_ssim(cpi->Source,
-                      &cm->post_proc_buffer, 1, &weight);
+                    frame_ssim2 = vpx_calc_ssim(cpi->Source,
+                      &cm->post_proc_buffer, &weight);
 
                     cpi->summed_quality += frame_ssim2 * weight;
                     cpi->summed_weights += weight;
@@ -5772,7 +5753,7 @@ int vp8_get_compressed_data(VP8_COMP *cpi, unsigned int *frame_flags, unsigned l
             if (cpi->b_calculate_ssimg)
             {
                 double y, u, v, frame_all;
-                frame_all =  vp8_calc_ssimg(cpi->Source, cm->frame_to_show,
+                frame_all = vpx_calc_ssimg(cpi->Source, cm->frame_to_show,
                     &y, &u, &v);
 
                 if (cpi->oxcf.number_of_layers > 1)
diff --git a/vp8/encoder/ssim.c b/vp8/encoder/ssim.c
deleted file mode 100644
index e75160836..000000000
--- a/vp8/encoder/ssim.c
+++ /dev/null
@@ -1,233 +0,0 @@
-/*
- *  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
- *
- *  Use of this source code is governed by a BSD-style license
- *  that can be found in the LICENSE file in the root of the source
- *  tree. An additional intellectual property rights grant can be found
- *  in the file PATENTS.  All contributing project authors may
- *  be found in the AUTHORS file in the root of the source tree.
- */
-
-
-#include "onyx_int.h"
-
-void vp8_ssim_parms_16x16_c
-(
-    unsigned char *s,
-    int sp,
-    unsigned char *r,
-    int rp,
-    unsigned long *sum_s,
-    unsigned long *sum_r,
-    unsigned long *sum_sq_s,
-    unsigned long *sum_sq_r,
-    unsigned long *sum_sxr
-)
-{
-    int i,j;
-    for(i=0;i<16;i++,s+=sp,r+=rp)
-     {
-         for(j=0;j<16;j++)
-         {
-             *sum_s += s[j];
-             *sum_r += r[j];
-             *sum_sq_s += s[j] * s[j];
-             *sum_sq_r += r[j] * r[j];
-             *sum_sxr += s[j] * r[j];
-         }
-     }
-}
-void vp8_ssim_parms_8x8_c
-(
-    unsigned char *s,
-    int sp,
-    unsigned char *r,
-    int rp,
-    unsigned long *sum_s,
-    unsigned long *sum_r,
-    unsigned long *sum_sq_s,
-    unsigned long *sum_sq_r,
-    unsigned long *sum_sxr
-)
-{
-    int i,j;
-    for(i=0;i<8;i++,s+=sp,r+=rp)
-     {
-         for(j=0;j<8;j++)
-         {
-             *sum_s += s[j];
-             *sum_r += r[j];
-             *sum_sq_s += s[j] * s[j];
-             *sum_sq_r += r[j] * r[j];
-             *sum_sxr += s[j] * r[j];
-         }
-     }
-}
-
-const static int64_t cc1 =  26634; // (64^2*(.01*255)^2
-const static int64_t cc2 = 239708; // (64^2*(.03*255)^2
-
-static double similarity
-(
-    unsigned long sum_s,
-    unsigned long sum_r,
-    unsigned long sum_sq_s,
-    unsigned long sum_sq_r,
-    unsigned long sum_sxr,
-    int count
-)
-{
-    int64_t ssim_n, ssim_d;
-    int64_t c1, c2;
-
-    //scale the constants by number of pixels
-    c1 = (cc1*count*count)>>12;
-    c2 = (cc2*count*count)>>12;
-
-    ssim_n = (2*sum_s*sum_r+ c1)*((int64_t) 2*count*sum_sxr-
-          (int64_t) 2*sum_s*sum_r+c2);
-
-    ssim_d = (sum_s*sum_s +sum_r*sum_r+c1)*
-        ((int64_t)count*sum_sq_s-(int64_t)sum_s*sum_s +
-        (int64_t)count*sum_sq_r-(int64_t) sum_r*sum_r +c2) ;
-
-    return ssim_n * 1.0 / ssim_d;
-}
-
-static double ssim_16x16(unsigned char *s,int sp, unsigned char *r,int rp)
-{
-    unsigned long sum_s=0,sum_r=0,sum_sq_s=0,sum_sq_r=0,sum_sxr=0;
-    vp8_ssim_parms_16x16(s, sp, r, rp, &sum_s, &sum_r, &sum_sq_s, &sum_sq_r, &sum_sxr);
-    return similarity(sum_s, sum_r, sum_sq_s, sum_sq_r, sum_sxr, 256);
-}
-static double ssim_8x8(unsigned char *s,int sp, unsigned char *r,int rp)
-{
-    unsigned long sum_s=0,sum_r=0,sum_sq_s=0,sum_sq_r=0,sum_sxr=0;
-    vp8_ssim_parms_8x8(s, sp, r, rp, &sum_s, &sum_r, &sum_sq_s, &sum_sq_r, &sum_sxr);
-    return similarity(sum_s, sum_r, sum_sq_s, sum_sq_r, sum_sxr, 64);
-}
-
-// TODO: (jbb) tried to scale this function such that we may be able to use it
-// for distortion metric in mode selection code ( provided we do a reconstruction)
-long dssim(unsigned char *s,int sp, unsigned char *r,int rp)
-{
-    unsigned long sum_s=0,sum_r=0,sum_sq_s=0,sum_sq_r=0,sum_sxr=0;
-    int64_t ssim3;
-    int64_t ssim_n1,ssim_n2;
-    int64_t ssim_d1,ssim_d2;
-    int64_t ssim_t1,ssim_t2;
-    int64_t c1, c2;
-
-    // normalize by 256/64
-    c1 = cc1*16;
-    c2 = cc2*16;
-
-    vp8_ssim_parms_16x16(s, sp, r, rp, &sum_s, &sum_r, &sum_sq_s, &sum_sq_r, &sum_sxr);
-    ssim_n1 = (2*sum_s*sum_r+ c1);
-
-    ssim_n2 =((int64_t) 2*256*sum_sxr-(int64_t) 2*sum_s*sum_r+c2);
-
-    ssim_d1 =((int64_t)sum_s*sum_s +(int64_t)sum_r*sum_r+c1);
-
-    ssim_d2 = (256 * (int64_t) sum_sq_s-(int64_t) sum_s*sum_s +
-                    (int64_t) 256*sum_sq_r-(int64_t) sum_r*sum_r +c2) ;
-
-    ssim_t1 = 256 - 256 * ssim_n1 / ssim_d1;
-    ssim_t2 = 256 - 256 * ssim_n2 / ssim_d2;
-
-    ssim3 = 256 *ssim_t1 * ssim_t2;
-    if(ssim3 <0 )
-        ssim3=0;
-    return (long)( ssim3  );
-}
-
-// We are using a 8x8 moving window with starting location of each 8x8 window
-// on the 4x4 pixel grid. Such arrangement allows the windows to overlap
-// block boundaries to penalize blocking artifacts.
-double vp8_ssim2
-(
-    unsigned char *img1,
-    unsigned char *img2,
-    int stride_img1,
-    int stride_img2,
-    int width,
-    int height
-)
-{
-    int i,j;
-    int samples =0;
-    double ssim_total=0;
-
-    // sample point start with each 4x4 location
-    for(i=0; i < height-8; i+=4, img1 += stride_img1*4, img2 += stride_img2*4)
-    {
-        for(j=0; j < width-8; j+=4 )
-        {
-            double v = ssim_8x8(img1+j, stride_img1, img2+j, stride_img2);
-            ssim_total += v;
-            samples++;
-        }
-    }
-    ssim_total /= samples;
-    return ssim_total;
-}
-double vp8_calc_ssim
-(
-    YV12_BUFFER_CONFIG *source,
-    YV12_BUFFER_CONFIG *dest,
-    int lumamask,
-    double *weight
-)
-{
-    double a, b, c;
-    double ssimv;
-
-    a = vp8_ssim2(source->y_buffer, dest->y_buffer,
-                 source->y_stride, dest->y_stride, source->y_width,
-                 source->y_height);
-
-    b = vp8_ssim2(source->u_buffer, dest->u_buffer,
-                 source->uv_stride, dest->uv_stride, source->uv_width,
-                 source->uv_height);
-
-    c = vp8_ssim2(source->v_buffer, dest->v_buffer,
-                 source->uv_stride, dest->uv_stride, source->uv_width,
-                 source->uv_height);
-
-    ssimv = a * .8 + .1 * (b + c);
-
-    *weight = 1;
-
-    return ssimv;
-}
-
-double vp8_calc_ssimg
-(
-    YV12_BUFFER_CONFIG *source,
-    YV12_BUFFER_CONFIG *dest,
-    double *ssim_y,
-    double *ssim_u,
-    double *ssim_v
-)
-{
-    double ssim_all = 0;
-    double a, b, c;
-
-    a = vp8_ssim2(source->y_buffer, dest->y_buffer,
-                 source->y_stride, dest->y_stride, source->y_width,
-                 source->y_height);
-
-    b = vp8_ssim2(source->u_buffer, dest->u_buffer,
-                 source->uv_stride, dest->uv_stride, source->uv_width,
-                 source->uv_height);
-
-    c = vp8_ssim2(source->v_buffer, dest->v_buffer,
-                 source->uv_stride, dest->uv_stride, source->uv_width,
-                 source->uv_height);
-    *ssim_y = a;
-    *ssim_u = b;
-    *ssim_v = c;
-    ssim_all = (a * 4 + b + c) /6;
-
-    return ssim_all;
-}
diff --git a/vp8/encoder/x86/ssim_opt_x86_64.asm b/vp8/encoder/x86/ssim_opt_x86_64.asm
deleted file mode 100644
index 5964a85f2..000000000
--- a/vp8/encoder/x86/ssim_opt_x86_64.asm
+++ /dev/null
@@ -1,216 +0,0 @@
-;
-;  Copyright (c) 2010 The WebM project authors. All Rights Reserved.
-;
-;  Use of this source code is governed by a BSD-style license
-;  that can be found in the LICENSE file in the root of the source
-;  tree. An additional intellectual property rights grant can be found
-;  in the file PATENTS.  All contributing project authors may
-;  be found in the AUTHORS file in the root of the source tree.
-;
-
-%include "vpx_ports/x86_abi_support.asm"
-
-; tabulate_ssim - sums sum_s,sum_r,sum_sq_s,sum_sq_r, sum_sxr
-%macro TABULATE_SSIM 0
-        paddusw         xmm15, xmm3  ; sum_s
-        paddusw         xmm14, xmm4  ; sum_r
-        movdqa          xmm1, xmm3
-        pmaddwd         xmm1, xmm1
-        paddd           xmm13, xmm1 ; sum_sq_s
-        movdqa          xmm2, xmm4
-        pmaddwd         xmm2, xmm2
-        paddd           xmm12, xmm2 ; sum_sq_r
-        pmaddwd         xmm3, xmm4
-        paddd           xmm11, xmm3  ; sum_sxr
-%endmacro
-
-; Sum across the register %1 starting with q words
-%macro SUM_ACROSS_Q 1
-        movdqa          xmm2,%1
-        punpckldq       %1,xmm0
-        punpckhdq       xmm2,xmm0
-        paddq           %1,xmm2
-        movdqa          xmm2,%1
-        punpcklqdq      %1,xmm0
-        punpckhqdq      xmm2,xmm0
-        paddq           %1,xmm2
-%endmacro
-
-; Sum across the register %1 starting with q words
-%macro SUM_ACROSS_W 1
-        movdqa          xmm1, %1
-        punpcklwd       %1,xmm0
-        punpckhwd       xmm1,xmm0
-        paddd           %1, xmm1
-        SUM_ACROSS_Q    %1
-%endmacro
-;void ssim_parms_sse2(
-;    unsigned char *s,
-;    int sp,
-;    unsigned char *r,
-;    int rp
-;    unsigned long *sum_s,
-;    unsigned long *sum_r,
-;    unsigned long *sum_sq_s,
-;    unsigned long *sum_sq_r,
-;    unsigned long *sum_sxr);
-;
-; TODO: Use parm passing through structure, probably don't need the pxors
-; ( calling app will initialize to 0 ) could easily fit everything in sse2
-; without too much hastle, and can probably do better estimates with psadw
-; or pavgb At this point this is just meant to be first pass for calculating
-; all the parms needed for 16x16 ssim so we can play with dssim as distortion
-; in mode selection code.
-global sym(vp8_ssim_parms_16x16_sse2) PRIVATE
-sym(vp8_ssim_parms_16x16_sse2):
-    push        rbp
-    mov         rbp, rsp
-    SHADOW_ARGS_TO_STACK 9
-    SAVE_XMM 15
-    push        rsi
-    push        rdi
-    ; end prolog
-
-    mov             rsi,        arg(0) ;s
-    mov             rcx,        arg(1) ;sp
-    mov             rdi,        arg(2) ;r
-    mov             rax,        arg(3) ;rp
-
-    pxor            xmm0, xmm0
-    pxor            xmm15,xmm15  ;sum_s
-    pxor            xmm14,xmm14  ;sum_r
-    pxor            xmm13,xmm13  ;sum_sq_s
-    pxor            xmm12,xmm12  ;sum_sq_r
-    pxor            xmm11,xmm11  ;sum_sxr
-
-    mov             rdx, 16      ;row counter
-.NextRow:
-
-    ;grab source and reference pixels
-    movdqu          xmm5, [rsi]
-    movdqu          xmm6, [rdi]
-    movdqa          xmm3, xmm5
-    movdqa          xmm4, xmm6
-    punpckhbw       xmm3, xmm0 ; high_s
-    punpckhbw       xmm4, xmm0 ; high_r
-
-    TABULATE_SSIM
-
-    movdqa          xmm3, xmm5
-    movdqa          xmm4, xmm6
-    punpcklbw       xmm3, xmm0 ; low_s
-    punpcklbw       xmm4, xmm0 ; low_r
-
-    TABULATE_SSIM
-
-    add             rsi, rcx   ; next s row
-    add             rdi, rax   ; next r row
-
-    dec             rdx        ; counter
-    jnz .NextRow
-
-    SUM_ACROSS_W    xmm15
-    SUM_ACROSS_W    xmm14
-    SUM_ACROSS_Q    xmm13
-    SUM_ACROSS_Q    xmm12
-    SUM_ACROSS_Q    xmm11
-
-    mov             rdi,arg(4)
-    movd            [rdi], xmm15;
-    mov             rdi,arg(5)
-    movd            [rdi], xmm14;
-    mov             rdi,arg(6)
-    movd            [rdi], xmm13;
-    mov             rdi,arg(7)
-    movd            [rdi], xmm12;
-    mov             rdi,arg(8)
-    movd            [rdi], xmm11;
-
-    ; begin epilog
-    pop         rdi
-    pop         rsi
-    RESTORE_XMM
-    UNSHADOW_ARGS
-    pop         rbp
-    ret
-
-;void ssim_parms_sse2(
-;    unsigned char *s,
-;    int sp,
-;    unsigned char *r,
-;    int rp
-;    unsigned long *sum_s,
-;    unsigned long *sum_r,
-;    unsigned long *sum_sq_s,
-;    unsigned long *sum_sq_r,
-;    unsigned long *sum_sxr);
-;
-; TODO: Use parm passing through structure, probably don't need the pxors
-; ( calling app will initialize to 0 ) could easily fit everything in sse2
-; without too much hastle, and can probably do better estimates with psadw
-; or pavgb At this point this is just meant to be first pass for calculating
-; all the parms needed for 16x16 ssim so we can play with dssim as distortion
-; in mode selection code.
-global sym(vp8_ssim_parms_8x8_sse2) PRIVATE
-sym(vp8_ssim_parms_8x8_sse2):
-    push        rbp
-    mov         rbp, rsp
-    SHADOW_ARGS_TO_STACK 9
-    SAVE_XMM 15
-    push        rsi
-    push        rdi
-    ; end prolog
-
-    mov             rsi,        arg(0) ;s
-    mov             rcx,        arg(1) ;sp
-    mov             rdi,        arg(2) ;r
-    mov             rax,        arg(3) ;rp
-
-    pxor            xmm0, xmm0
-    pxor            xmm15,xmm15  ;sum_s
-    pxor            xmm14,xmm14  ;sum_r
-    pxor            xmm13,xmm13  ;sum_sq_s
-    pxor            xmm12,xmm12  ;sum_sq_r
-    pxor            xmm11,xmm11  ;sum_sxr
-
-    mov             rdx, 8      ;row counter
-.NextRow:
-
-    ;grab source and reference pixels
-    movq            xmm3, [rsi]
-    movq            xmm4, [rdi]
-    punpcklbw       xmm3, xmm0 ; low_s
-    punpcklbw       xmm4, xmm0 ; low_r
-
-    TABULATE_SSIM
-
-    add             rsi, rcx   ; next s row
-    add             rdi, rax   ; next r row
-
-    dec             rdx        ; counter
-    jnz .NextRow
-
-    SUM_ACROSS_W    xmm15
-    SUM_ACROSS_W    xmm14
-    SUM_ACROSS_Q    xmm13
-    SUM_ACROSS_Q    xmm12
-    SUM_ACROSS_Q    xmm11
-
-    mov             rdi,arg(4)
-    movd            [rdi], xmm15;
-    mov             rdi,arg(5)
-    movd            [rdi], xmm14;
-    mov             rdi,arg(6)
-    movd            [rdi], xmm13;
-    mov             rdi,arg(7)
-    movd            [rdi], xmm12;
-    mov             rdi,arg(8)
-    movd            [rdi], xmm11;
-
-    ; begin epilog
-    pop         rdi
-    pop         rsi
-    RESTORE_XMM
-    UNSHADOW_ARGS
-    pop         rbp
-    ret
diff --git a/vp8/vp8cx.mk b/vp8/vp8cx.mk
index 59f9fe839..ea7d472ae 100644
--- a/vp8/vp8cx.mk
+++ b/vp8/vp8cx.mk
@@ -65,7 +65,6 @@ VP8_CX_SRCS-yes += encoder/ratectrl.c
 VP8_CX_SRCS-yes += encoder/rdopt.c
 VP8_CX_SRCS-yes += encoder/segmentation.c
 VP8_CX_SRCS-yes += encoder/segmentation.h
-VP8_CX_SRCS-$(CONFIG_INTERNAL_STATS) += encoder/ssim.c
 VP8_CX_SRCS-yes += encoder/tokenize.c
 VP8_CX_SRCS-yes += encoder/dct_value_cost.h
 VP8_CX_SRCS-yes += encoder/dct_value_tokens.h
@@ -97,7 +96,6 @@ VP8_CX_SRCS-$(HAVE_SSE2) += encoder/x86/temporal_filter_apply_sse2.asm
 VP8_CX_SRCS-$(HAVE_SSE2) += encoder/x86/vp8_enc_stubs_sse2.c
 VP8_CX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += encoder/x86/quantize_mmx.asm
 VP8_CX_SRCS-$(ARCH_X86)$(ARCH_X86_64) += encoder/x86/encodeopt.asm
-VP8_CX_SRCS-$(ARCH_X86_64) += encoder/x86/ssim_opt_x86_64.asm
 
 ifeq ($(CONFIG_REALTIME_ONLY),yes)
 VP8_CX_SRCS_REMOVE-$(HAVE_SSE2) += encoder/x86/temporal_filter_apply_sse2.asm