Open main menu

CDOT Wiki β

Difference between revisions of "Three-Star"

(LZW Data Compression Algorithm)
(Assignment 3)
 
(63 intermediate revisions by 2 users not shown)
Line 10: Line 10:
 
=== Assignment 1 ===
 
=== Assignment 1 ===
  
==== Image Profiling ====
+
==== Image Processor Application ====
Chosen to profile image profiling as shown here: http://www.dreamincode.net/forums/topic/76816-image-processing-tutorial/ , using the sample program files
+
Chosen to profile image processing as shown here: http://www.dreamincode.net/forums/topic/76816-image-processing-tutorial/ , using the sample program files
(main/image.h/image.cpp)
+
(main/image.h/image.cpp), where it processeses PGM(portable gray map) files.
 +
 
  
 
pulled PGM sample files from here: https://userpages.umbc.edu/~rostamia/2003-09-math625/images.html
 
pulled PGM sample files from here: https://userpages.umbc.edu/~rostamia/2003-09-math625/images.html
Line 35: Line 36:
  
 
[[File:Callgraphpt1.png]]
 
[[File:Callgraphpt1.png]]
 
[[File:Callgraphpt2.png]]
 
  
 
Rotate image function is one of the longer running functions and looks like it has potential for parallelization.
 
Rotate image function is one of the longer running functions and looks like it has potential for parallelization.
Line 167: Line 166:
  
 
     for (int s = 1; s <= n/2; s*=2)
 
     for (int s = 1; s <= n/2; s*=2)
        for(int j = 0; j < n; j +=2 * s)
+
      for(int j = 0; j < n; j +=2 * s)
    a[j] += a[j + s];
+
        a[j] += a[j + s];
  
 
As such, the major hotspot in this function is the second for loop. This is especially true since the file might be very large and we may be dealing with millions of characters! The one thing we need to worry about is that order does seem to matter for the second for loop.
 
As such, the major hotspot in this function is the second for loop. This is especially true since the file might be very large and we may be dealing with millions of characters! The one thing we need to worry about is that order does seem to matter for the second for loop.
  
 
==== Conclusion ====
 
==== Conclusion ====
We decided to go with image profiling.  
+
We decided to go with image profiling. It is a pretty simple parallelization since the transformation functions are matrix transformations which don't care about which element is processed first.
  
 
There are some possible issues with working with the simple-lzw-compression-algorithm and CUDA. You cannot use the C++ string type in a kernel because CUDA does not include a device version of the C++ String library that would be able run on the GPU. Even if it was possible to use string in a kernel, it's not something you would want to do because string handles memory dynamically, which would be likely to be slow.
 
There are some possible issues with working with the simple-lzw-compression-algorithm and CUDA. You cannot use the C++ string type in a kernel because CUDA does not include a device version of the C++ String library that would be able run on the GPU. Even if it was possible to use string in a kernel, it's not something you would want to do because string handles memory dynamically, which would be likely to be slow.
  
Char explanation (replace tmr) - https://stackoverflow.com/questions/26993351/is-there-a-penalty-to-using-char-variables-in-cuda-kernels?utm_medium=organic&utm_source=google_rich_qa&utm_campaign=google_rich_qa
+
Essentially, using chars on the gpu would require we use 8-bit arithmetic and need to convert from 32-bit arithmetic for operations. On top of that, the compress function refers to a map of strings/int pairs to shrink the size of the file. Even if we did manage the character operations, it would need to somehow use the string map to get the corresponding integer which could lead to being unable to use device memory for optimization.
  
 
=== Assignment 2 ===
 
=== Assignment 2 ===
 +
 +
 +
Original CPU Implementation:
 +
 +
  void Image::rotateImage(int theta, Image& oldImage)
 +
  /*based on users input and rotates it around the center of the image.*/
 +
  {
 +
    int r0, c0;
 +
    int r1, c1;
 +
    int rows, cols;
 +
    rows = oldImage.N;
 +
    cols = oldImage.M;
 +
    Image tempImage(rows, cols, oldImage.Q);
 +
   
 +
    float rads = (theta * 3.14159265)/180.0;
 +
   
 +
    r0 = rows / 2;
 +
    c0 = cols / 2;
 +
   
 +
    for(int r = 0; r < rows; r++)
 +
    {
 +
        for(int c = 0; c < cols; c++)
 +
        {
 +
            r1 = (int) (r0 + ((r - r0) * cos(rads)) - ((c - c0) * sin(rads)));
 +
            c1 = (int) (c0 + ((r - r0) * sin(rads)) + ((c - c0) * cos(rads)));
 +
           
 +
            if(inBounds(r1,c1))
 +
            {
 +
                tempImage.pixelVal[r1][c1] = oldImage.pixelVal[r][c];
 +
            }
 +
        }
 +
    }
 +
   
 +
    for(int i = 0; i < rows; i++)
 +
    {
 +
        for(int j = 0; j < cols; j++)
 +
        {
 +
            if(tempImage.pixelVal[i][j] == 0)
 +
                tempImage.pixelVal[i][j] = tempImage.pixelVal[i][j+1];
 +
        }
 +
    }
 +
    oldImage = tempImage;
 +
  }
 +
 +
Parallelized Code (done by Timothy Moy, Derrick acted as consulting for how to use the program):
 +
 +
Kernels
 +
 +
  __device__ bool inBounds(int row, int col, int maxRow, int maxCol) {
 +
    if (row >= maxRow || row < 0 || col >= maxCol || col < 0)
 +
      return false;
 +
      //else
 +
      return true;
 +
  }
 +
   
 +
  __global__ void rotateKernel(int* oldImage, int* newImage, int rows, int cols, float rads) {
 +
    int r = blockIdx.x * blockDim.x + threadIdx.x;
 +
    int c = blockIdx.y * blockDim.y + threadIdx.y;
 +
   
 +
    int r0 = rows / 2;
 +
    int c0 = cols / 2;
 +
    float sinRads = sinf(rads);
 +
    float cosRads = cosf(rads);
 +
   
 +
    /*__shared__ int s[ntpb * ntpb];
 +
    s[r * cols + c] = oldImage[r * cols + c];*/
 +
   
 +
    if (r < rows && c < cols)
 +
    {
 +
      int r1 = (int)(r0 + ((r - r0) * cosRads) - ((c - c0) * sinRads));
 +
      int c1 = (int)(c0 + ((r - r0) * sinRads) + ((c - c0) * cosRads));
 +
   
 +
      if (inBounds(r1, c1, rows, cols))
 +
      {
 +
        newImage[r1 * cols + c1] = oldImage[r * cols + c];
 +
      }
 +
    }
 +
  }
 +
 +
Modified Function
 +
 +
  void Image::rotateImage(int theta, Image& oldImage)
 +
  /*based on users input and rotates it around the center of the image.*/
 +
  {
 +
    int r0, c0;
 +
    int r1, c1;
 +
    int rows, cols;
 +
    rows = oldImage.N;
 +
    cols = oldImage.M;
 +
    Image tempImage(rows, cols, oldImage.Q);
 +
   
 +
    float rads = (theta * 3.14159265)/180.0;
 +
   
 +
    // workspace start
 +
    // - calculate number of blocks for n rows assume square image
 +
    int nb = (rows + ntpb - 1) / ntpb;
 +
   
 +
    // allocate memory for matrices d_a, d_b on the device
 +
 
 +
    // - add your allocation code here
 +
    int* d_a;
 +
    check("device a", cudaMalloc((void**)&d_a, rows* cols * sizeof(int)));
 +
    int* d_b;
 +
    check("device b", cudaMalloc((void**)&d_b, rows* cols * sizeof(int)));
 +
   
 +
    // copy h_a and h_b to d_a and d_b (host to device)
 +
    // - add your copy code here
 +
    check("copy to d_a", cudaMemcpy(d_a, oldImage.pixelVal, rows * cols * sizeof(int), cudaMemcpyHostToDevice));
 +
    //check("copy to d_b", cudaMemcpy(d_b, tempImage.pixelVal, rows * cols * sizeof(int), cudaMemcpyHostToDevice));
 +
   
 +
    // launch execution configuration
 +
    // - define your 2D grid of blocks
 +
    dim3 dGrid(nb, nb);
 +
    // - define your 2D block of threads
 +
    dim3 dBlock(ntpb, ntpb);
 +
    // - launch your execution configuration
 +
    rotateKernel<<<dGrid, dBlock >>>(d_a, d_b, rows, cols, rads);
 +
    check("launch error: ", cudaPeekAtLastError());
 +
    // - check for launch termination
 +
    // synchronize the device and the host
 +
    check("Synchronize ", cudaDeviceSynchronize());
 +
   
 +
    // copy d_b to tempImage (device to host)
 +
    // - add your copy code here
 +
    check("device copy to hc", cudaMemcpy(tempImage.pixelVal, d_b, rows * cols * sizeof(int), cudaMemcpyDeviceToHost));
 +
   
 +
    // deallocate device memory
 +
    // - add your deallocation code here
 +
    cudaFree(d_a);
 +
    cudaFree(d_b);
 +
   
 +
    // reset the device
 +
    cudaDeviceReset();
 +
    // workspace end
 +
   
 +
    for(int i = 0; i < rows; i++)
 +
    {
 +
        for(int j = 0; j < cols; j++)
 +
        {
 +
            if(tempImage.pixelVal[i * M + j] == 0)
 +
                tempImage.pixelVal[i * M + j] = tempImage.pixelVal[i * M + j+1];
 +
        }
 +
    }
 +
    oldImage = tempImage;
 +
  }
 +
 +
Profiling (Done by Derrick Leung)
 +
 +
{|
 +
|'''Function'''
 +
|'''CPU-Only'''
 +
|'''GPU-CPU'''
 +
|'''speedup(%)'''
 +
|-
 +
|'''Cuda Memory Allocation'''
 +
|n/a
 +
| 1164 ms
 +
|n/a
 +
|-
 +
|'''Copy Image to Device memory'''
 +
|n/a
 +
| 6 ms
 +
|n/a
 +
|-
 +
|'''Kernel'''
 +
|n/a
 +
| 0 ms
 +
|n/a
 +
|-
 +
|'''Copy device image to host temp variable'''
 +
|n/a
 +
| 6 ms
 +
|n/a
 +
|-
 +
|'''copy temp image to original image variable'''
 +
|n/a
 +
| 43 ms
 +
|n/a
 +
|-
 +
|'''Total Rotation Time (no allocation, with memcpy)'''
 +
|1717ms
 +
| 55ms
 +
| 3021.82%
 +
|-
 +
|'''Total Run Time'''
 +
|1775 ms
 +
|1294 ms
 +
| 37.17%
 +
|}
 +
 +
Comparisons
 +
 +
{|
 +
|
 +
!Rotation Run Time (exclude memory allocation)||||
 +
!Total Run Time
 +
|-
 +
!Size of Picture
 +
!CPU-Only
 +
!GPU-CPU
 +
!speedup ratio
 +
!CPU-Only
 +
!GPU-CPU
 +
!speedup ratio
 +
|-
 +
|512x512
 +
| 67 ms
 +
| 2ms
 +
| 33.50
 +
| 71 ms
 +
| 372 ms
 +
| .19
 +
|-
 +
|2x enlarged
 +
| 265 ms
 +
| 7 ms
 +
| 37.85
 +
| 277 ms
 +
| 410 ms
 +
| .67
 +
|-
 +
|3x enlarged
 +
| 608 ms
 +
| 23 ms
 +
| 26.43
 +
| 630 ms
 +
| 427 ms
 +
| 1.47
 +
|-
 +
|4x enlarged
 +
| 1091 ms
 +
| 37 ms
 +
| 29.48
 +
| 1129 ms
 +
| 446 ms
 +
| 2.53
 +
|-
 +
|5x enlarged
 +
| 1717 ms
 +
| 55 ms
 +
| 31.22
 +
| 1775 ms
 +
| 476 ms
 +
| 3.73
 +
|-
 +
|}
 +
 +
[[File:assign2-rotate-comparison.png]]
 +
[[File:assign2-total-comparison.png]]
 +
 +
Excel Sheet
 +
[[File:assignment2 profile.xlsx.txt]]
 +
 +
Source Code:
 +
[[File:image.cu.txt]]
 +
[[File:image.h.txt]]
 +
[[File:main.cpp.txt]]
  
 
=== Assignment 3 ===
 
=== Assignment 3 ===
 +
 +
'''Shared Memory''' (Derrick Leung)
 +
 +
Shared memory does not really help, because we are not performing any computations on the matrix in the kernel - only thing being done is copying memory. 
 +
 +
'''Coalesced Memory''' (Derrick Leung)
 +
 +
changed matrix access from column to row(16x16 block size)
 +
 +
[[File:Coaslescedchangepng.png]]
 +
 +
{|
 +
|
 +
!Uncoalesced ||||
 +
!Coalesced
 +
|-
 +
!Size of Picture
 +
!memcpy
 +
!rotate kernel
 +
!total runtime
 +
!memcpy
 +
!rotate kernel
 +
!total runtime
 +
 +
|-
 +
|512x512
 +
| 0.50ms
 +
| 0.89ms
 +
| 91.77ms
 +
| 0.51ms
 +
| 0.89ms
 +
| 97.13ms
 +
|-
 +
|2x enlarged
 +
| 1.91ms
 +
| 3.56ms
 +
| 97.78ms
 +
| 1.81ms
 +
| 3.54ms
 +
| 98.82ms
 +
|-
 +
|3x enlarged
 +
| 4.63ms
 +
| 7.97ms
 +
| 107.07ms
 +
| 4.38ms
 +
| 7.95ms
 +
| 105.41ms
 +
|-
 +
|4x enlarged
 +
| 7.71ms
 +
| 14.15ms
 +
| 112.54ms
 +
| 7.61ms
 +
| 14.10ms
 +
| 111.86ms
 +
|-
 +
|5x enlarged
 +
| 12.80ms
 +
| 22.10ms
 +
| 131.18ms
 +
| 12.60ms
 +
| 22.00ms
 +
| 126.04ms
 +
|-
 +
|}
 +
 +
[[File:Uncoalesced vs coalesced excel chart.png]]
 +
 +
Changing the way memory is accessed doesn't seem to have any significant improvements/changes to time
 +
 +
 +
'''Tiling''' (Timothy Moy)
 +
 +
Tiling ended up being a no go as we didn't even have a use of implementing the shared memory. Since we weren't using shared memory, and tiling improves performance via shared memory we opted not to try implementing it and try other methods instead.
 +
 +
'''Block Size''' (Timothy Moy)
 +
 +
The code modified was line 22
 +
 +
    const int ntpb = 16; // number of threads per block
 +
 +
The first quick method to try and improve it was to change the block size. Playing with the block size changed the kernel run times, but it wasn't apparent what exactly causes it. Most likely it is due to the 16*16 block configuration being able to not take up all the memory of the SM, but is still large enough that it gives us a boost in execution times. https://devtalk.nvidia.com/default/topic/1026825/how-to-choose-how-many-threads-blocks-to-have-/
 +
 +
[[Media:Assign3-ntpb.png]]
 +
 +
In the end, a block size of 16 by 16 proved to be best for run times.
 +
 +
'''Moving Repeated Calculations to the Host''' (Timothy Moy)
 +
 +
I then tried merging the sinf() and cosf() function calls into one via sincosf() so that the kernel made less function calls. That proved to be trim the run times a bit, but then I noticed that sin and cos never change since our angle never changes. Thus, this led to testing of the sin and cos functions to use the Host to calculate it and pass them in as parameters for the kernel. The result was a much more significant run time since our kernel is no longer calculating the same number in each thread.
 +
 +
Kernel Signature Changes:
 +
 +
    __global__ void rotateKernel(int* oldImage, int* newImage, int rows, int cols, float rads) {
 +
 +
vs
 +
 +
    __global__ void rotateKernel(int* oldImage, int* newImage, int rows, int cols, /*float rads*/ float sinRads, float cosRads) {
 +
 +
Kernel Code  Changes
 +
 +
    float sinRads = sinf(rads);
 +
    float cosRads = cosf(rads);
 +
   
 +
    //float sinRads, cosRads;
 +
    //__sincosf(rads, &sinRads, &cosRads);
 +
 +
vs
 +
 +
    //float sinRads = sinf(rads);
 +
    //float cosRads = cosf(rads);
 +
   
 +
    float sinRads, cosRads;
 +
    __sincosf(rads, &sinRads, &cosRads);
 +
 +
vs
 +
 +
    //float sinRads = sinf(rads);
 +
    //float cosRads = cosf(rads);
 +
   
 +
    //float sinRads, cosRads;
 +
    //__sincosf(rads, &sinRads, &cosRads);
 +
 +
and
 +
 +
Host Function Additions:
 +
 +
    float cos1 = cos(rads);
 +
    float sin1 = sin(rads);
 +
 +
Kernel Launch Changed
 +
 +
    rotateKernel<<<dGrid, dBlock >>>(d_a, d_b, rows, cols, rads);
 +
 +
vs
 +
 +
    rotateKernel<<<dGrid, dBlock >>>(d_a, d_b, rows, cols, sin1, cos1);
 +
 +
The graph below shows the pronounced difference between the different sin cos methods.
 +
 +
[[File:assign3-all.png]]
 +
 +
There may be other variables that could be moved outside the kernel like r0 and c0, but due to time limitations they weren't tested.
 +
 +
All assignments compared in this file under "a1a2a3comps" sheet.
 +
 +
[[File:Assignment3_profile.xlsx.txt]]
 +
 +
 +
'''Github'''
 +
 +
https://github.com/dleung25/GPU610-Assignment3-Image-Profiling
 +
 +
[We used the images in the Github and 100 degrees for all our tests]

Latest revision as of 08:54, 9 April 2018

Three-Star

Team Members

  1. Derrick Leung
  2. Timothy Moy

Email All

Progress

Assignment 1

Image Processor Application

Chosen to profile image processing as shown here: http://www.dreamincode.net/forums/topic/76816-image-processing-tutorial/ , using the sample program files (main/image.h/image.cpp), where it processeses PGM(portable gray map) files.


pulled PGM sample files from here: https://userpages.umbc.edu/~rostamia/2003-09-math625/images.html

file sizes being 512x512, about 262 KB each file

Compiled to produce a flat profile and a call graph

>g++ -g -O2 -pg -o main main.cpp

>main a.pgm result.pgm

Note: Enlarged image by max permitted by program (5) to get more viewable results, since the profile without enlarging it produces non-significant results

 

The results of the flat profile:

 

The results of the call graph

 

Rotate image function is one of the longer running functions and looks like it has potential for parallelization.

 

LZW Data Compression Algorithm

Timothy Moy profiled.

Original algorithm: https://codereview.stackexchange.com/questions/86543/simple-lzw-compression-algorithm

Raw Flat profile (50Mb Test file for compression):

Each sample counts as 0.01 seconds.

 %   cumulative   self              self     total           
time   seconds   seconds    calls  us/call  us/call  name    
35.52      4.23     4.23                             compress(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >)
27.54      7.51     3.28 102062309     0.03     0.03  std::_Hashtable<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::pair<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, int>, std::allocator<std::pair<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, int> >, std::__detail::_Select1st, std::equal_to<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::hash<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::__detail::_Mod_range_hashing, std::__detail::_Default_ranged_hash, std::__detail::_Prime_rehash_policy, std::__detail::_Hashtable_traits<true, false, true> >::_M_find_before_node(unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&, unsigned int) const
20.15      9.91     2.40 204116423     0.01     0.01  std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_replace_aux(unsigned int, unsigned int, unsigned int, char)
 8.23     10.89     0.98 49629412     0.02     0.05  std::__detail::_Map_base<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::pair<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, int>, std::allocator<std::pair<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, int> >, std::__detail::_Select1st, std::equal_to<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::hash<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::__detail::_Mod_range_hashing, std::__detail::_Default_ranged_hash, std::__detail::_Prime_rehash_policy, std::__detail::_Hashtable_traits<true, false, true>, true>::operator[](std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)
 4.28     11.40     0.51 52428800     0.01     0.01  std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_assign(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)
 3.02     11.76     0.36 52436762     0.01     0.01  show_usage()
 1.26     11.91     0.15                             _Z22convert_char_to_stringB5cxx11PKci
 0.00     11.91     0.00     4097     0.00     0.00  std::_Hashtable<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::pair<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, int>, std::allocator<std::pair<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, int> >, std::__detail::_Select1st, std::equal_to<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::hash<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::__detail::_Mod_range_hashing, std::__detail::_Default_ranged_hash, std::__detail::_Prime_rehash_policy, std::__detail::_Hashtable_traits<true, false, true> >::_M_insert_unique_node(unsigned int, unsigned int, std::__detail::_Hash_node<std::pair<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const, int>, true>*)
 0.00     11.91     0.00       22     0.00     0.01  std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >::_M_mutate(unsigned int, unsigned int, char const*, unsigned int)
 0.00     11.91     0.00        1     0.00     0.00  _GLOBAL__sub_I__Z18convert_int_to_binB5cxx11i
 0.00     11.91     0.00        1     0.00    28.13  std::_Hashtable<std::__cxx11::basic_

Summarized Flat Profile (50Mb Test file for compression):

 %   cumulative   self              self     total           
time   seconds   seconds    calls  us/call  us/call  name    
35.52      4.23     4.23                             compress()
27.54      7.51     3.28 102062309    0.03     0.03  std::_Hashtable
20.15      9.91     2.40 204116423    0.01     0.01  std::__cxx11::basic_string
 8.23     10.89     0.98 49629412     0.02     0.05  std::__detail::_Map_base
 4.28     11.40     0.51 52428800     0.01     0.01  std::__cxx11::basic_string
 3.02     11.76     0.36 52436762     0.01     0.01  show_usage()
 1.26     11.91     0.15                             _Z22convert_char_to_stringB5cxx11PKci
 0.00     11.91     0.00     4097     0.00     0.00  std::_Hashtable
 0.00     11.91     0.00       22     0.00     0.01  std::__cxx11::basic_string
 0.00     11.91     0.00        1     0.00     0.00  _GLOBAL__sub_I__Z18convert_int_to_binB5cxx11i
 0.00     11.91     0.00        1     0.00    28.13  std::_Hashtable<std::__cxx11::basic_

Note how the compress() function takes up the largest amount of time (over one third), then the other functions which take up over 10% of the time are library functions. It is highly unlikely we could parallelize the library functions. The other functions that take up under 10% of the time will probably not give enough improvement in time to make a significant impact.

Thus, the function we should focus on is the compress function.

Summary of Compress() Profiles

Size (MB) Compress() time in seconds
10 0.96
15 1.35
20 1.8
25 2.14
30 2.64
35 3.16
40 3.45
45 4.24
50 4.23

The compress function source code:

void compress(string input, int size, string filename) {

   unordered_map<string, int> compress_dictionary(MAX_DEF);
   //Dictionary initializing with ASCII
   for ( int unsigned i = 0 ; i < 256 ; i++ ){
   compress_dictionary[string(1,i)] = i;
   }
   string current_string;
   unsigned int code;
   unsigned int next_code = 256;
   //Output file for compressed data
   ofstream outputFile;
   outputFile.open(filename + ".lzw");
   // Possible area for improvement via reduction
   for(char& c: input){
   current_string = current_string + c;
   if ( compress_dictionary.find(current_string) ==compress_dictionary.end() ){
           if (next_code <= MAX_DEF)
               compress_dictionary.insert(make_pair(current_string, next_code++));
           current_string.erase(current_string.size()-1);
           outputFile << convert_int_to_bin(compress_dictionary[current_string]);
           current_string = c;
       }   
   }   
   if (current_string.size())
           outputFile << convert_int_to_bin(compress_dictionary[current_string]);
   outputFile.close();

}

There are two loops which show possibility of parallelization:

   for ( int unsigned i = 0 ; i < 256 ; i++ ){
       compress_dictionary[string(1,i)] = i;
   }

and

   for(char& c: input){
       current_string = current_string + c; // Possible area for improvement via reduction
       if ( compress_dictionary.find(current_string) ==compress_dictionary.end() ){
           if (next_code <= MAX_DEF)
               compress_dictionary.insert(make_pair(current_string, next_code++));
           current_string.erase(current_string.size()-1);
           outputFile << convert_int_to_bin(compress_dictionary[current_string]);
           current_string = c;
       }   
   }   

The first for loop is constant and probably won't show much improvement if we parallelize it.


Note the comment above the second for loop notes we can do something like this:

   for (int i = 1; i < n; i+=) a[0] += a[i];

changed to

   for (int s = 1; s <= n/2; s*=2)
     for(int j = 0; j < n; j +=2 * s)
       a[j] += a[j + s];

As such, the major hotspot in this function is the second for loop. This is especially true since the file might be very large and we may be dealing with millions of characters! The one thing we need to worry about is that order does seem to matter for the second for loop.

Conclusion

We decided to go with image profiling. It is a pretty simple parallelization since the transformation functions are matrix transformations which don't care about which element is processed first.

There are some possible issues with working with the simple-lzw-compression-algorithm and CUDA. You cannot use the C++ string type in a kernel because CUDA does not include a device version of the C++ String library that would be able run on the GPU. Even if it was possible to use string in a kernel, it's not something you would want to do because string handles memory dynamically, which would be likely to be slow.

Essentially, using chars on the gpu would require we use 8-bit arithmetic and need to convert from 32-bit arithmetic for operations. On top of that, the compress function refers to a map of strings/int pairs to shrink the size of the file. Even if we did manage the character operations, it would need to somehow use the string map to get the corresponding integer which could lead to being unable to use device memory for optimization.

Assignment 2

Original CPU Implementation:

 void Image::rotateImage(int theta, Image& oldImage)
 /*based on users input and rotates it around the center of the image.*/
 {
   int r0, c0;
   int r1, c1;
   int rows, cols;
   rows = oldImage.N;
   cols = oldImage.M;
   Image tempImage(rows, cols, oldImage.Q);
   
   float rads = (theta * 3.14159265)/180.0;
   
   r0 = rows / 2;
   c0 = cols / 2;
   
   for(int r = 0; r < rows; r++)
   {
       for(int c = 0; c < cols; c++)
       {
           r1 = (int) (r0 + ((r - r0) * cos(rads)) - ((c - c0) * sin(rads)));
           c1 = (int) (c0 + ((r - r0) * sin(rads)) + ((c - c0) * cos(rads)));
           
           if(inBounds(r1,c1))
           {
               tempImage.pixelVal[r1][c1] = oldImage.pixelVal[r][c];
           }
       }
   }
   
   for(int i = 0; i < rows; i++)
   {
       for(int j = 0; j < cols; j++)
       {
           if(tempImage.pixelVal[i][j] == 0)
               tempImage.pixelVal[i][j] = tempImage.pixelVal[i][j+1];
       }
   }
   oldImage = tempImage;
 }

Parallelized Code (done by Timothy Moy, Derrick acted as consulting for how to use the program):

Kernels

 __device__ bool inBounds(int row, int col, int maxRow, int maxCol) {
   if (row >= maxRow || row < 0 || col >= maxCol || col < 0)
     return false;
     //else
     return true;
 }
   
 __global__ void rotateKernel(int* oldImage, int* newImage, int rows, int cols, float rads) {
   int r = blockIdx.x * blockDim.x + threadIdx.x;
   int c = blockIdx.y * blockDim.y + threadIdx.y;
   
   int r0 = rows / 2;
   int c0 = cols / 2;
   float sinRads = sinf(rads);
   float cosRads = cosf(rads);
   
   /*__shared__ int s[ntpb * ntpb];
   s[r * cols + c] = oldImage[r * cols + c];*/
   
   if (r < rows && c < cols)
   {
     int r1 = (int)(r0 + ((r - r0) * cosRads) - ((c - c0) * sinRads));
     int c1 = (int)(c0 + ((r - r0) * sinRads) + ((c - c0) * cosRads));
   
     if (inBounds(r1, c1, rows, cols))
     {
       newImage[r1 * cols + c1] = oldImage[r * cols + c];
     }
   }
 }

Modified Function

 void Image::rotateImage(int theta, Image& oldImage)
 /*based on users input and rotates it around the center of the image.*/
 {
   int r0, c0;
   int r1, c1;
   int rows, cols;
   rows = oldImage.N;
   cols = oldImage.M;
   Image tempImage(rows, cols, oldImage.Q);
   
   float rads = (theta * 3.14159265)/180.0;
   
   // workspace start
   // - calculate number of blocks for n rows assume square image
   int nb = (rows + ntpb - 1) / ntpb;
   
   // allocate memory for matrices d_a, d_b on the device
 
   // - add your allocation code here
   int* d_a;
   check("device a", cudaMalloc((void**)&d_a, rows* cols * sizeof(int)));
   int* d_b;
   check("device b", cudaMalloc((void**)&d_b, rows* cols * sizeof(int)));
   
   // copy h_a and h_b to d_a and d_b (host to device)
   // - add your copy code here
   check("copy to d_a", cudaMemcpy(d_a, oldImage.pixelVal, rows * cols * sizeof(int), cudaMemcpyHostToDevice));
   //check("copy to d_b", cudaMemcpy(d_b, tempImage.pixelVal, rows * cols * sizeof(int), cudaMemcpyHostToDevice));
   
   // launch execution configuration
   // - define your 2D grid of blocks
   dim3 dGrid(nb, nb);
   // - define your 2D block of threads
   dim3 dBlock(ntpb, ntpb);
   // - launch your execution configuration	
   rotateKernel<<<dGrid, dBlock >>>(d_a, d_b, rows, cols, rads);
   check("launch error: ", cudaPeekAtLastError());
   // - check for launch termination
   // synchronize the device and the host
   check("Synchronize ", cudaDeviceSynchronize());
   
   // copy d_b to tempImage (device to host)
   // - add your copy code here
   check("device copy to hc", cudaMemcpy(tempImage.pixelVal, d_b, rows * cols * sizeof(int), cudaMemcpyDeviceToHost));
   
   // deallocate device memory
   // - add your deallocation code here
   cudaFree(d_a);
   cudaFree(d_b);
   
   // reset the device
   cudaDeviceReset();
   // workspace end
   
   for(int i = 0; i < rows; i++)
   {
       for(int j = 0; j < cols; j++)
       {
           if(tempImage.pixelVal[i * M + j] == 0)
               tempImage.pixelVal[i * M + j] = tempImage.pixelVal[i * M + j+1];
       }
   }
   oldImage = tempImage;
 }

Profiling (Done by Derrick Leung)

Function CPU-Only GPU-CPU speedup(%)
Cuda Memory Allocation n/a 1164 ms n/a
Copy Image to Device memory n/a 6 ms n/a
Kernel n/a 0 ms n/a
Copy device image to host temp variable n/a 6 ms n/a
copy temp image to original image variable n/a 43 ms n/a
Total Rotation Time (no allocation, with memcpy) 1717ms 55ms 3021.82%
Total Run Time 1775 ms 1294 ms 37.17%

Comparisons

Rotation Run Time (exclude memory allocation) Total Run Time
Size of Picture CPU-Only GPU-CPU speedup ratio CPU-Only GPU-CPU speedup ratio
512x512 67 ms 2ms 33.50 71 ms 372 ms .19
2x enlarged 265 ms 7 ms 37.85 277 ms 410 ms .67
3x enlarged 608 ms 23 ms 26.43 630 ms 427 ms 1.47
4x enlarged 1091 ms 37 ms 29.48 1129 ms 446 ms 2.53
5x enlarged 1717 ms 55 ms 31.22 1775 ms 476 ms 3.73

   

Excel Sheet File:Assignment2 profile.xlsx.txt

Source Code: File:Image.cu.txt File:Image.h.txt File:Main.cpp.txt

Assignment 3

Shared Memory (Derrick Leung)

Shared memory does not really help, because we are not performing any computations on the matrix in the kernel - only thing being done is copying memory.

Coalesced Memory (Derrick Leung)

changed matrix access from column to row(16x16 block size)

 

Uncoalesced Coalesced
Size of Picture memcpy rotate kernel total runtime memcpy rotate kernel total runtime
512x512 0.50ms 0.89ms 91.77ms 0.51ms 0.89ms 97.13ms
2x enlarged 1.91ms 3.56ms 97.78ms 1.81ms 3.54ms 98.82ms
3x enlarged 4.63ms 7.97ms 107.07ms 4.38ms 7.95ms 105.41ms
4x enlarged 7.71ms 14.15ms 112.54ms 7.61ms 14.10ms 111.86ms
5x enlarged 12.80ms 22.10ms 131.18ms 12.60ms 22.00ms 126.04ms

 

Changing the way memory is accessed doesn't seem to have any significant improvements/changes to time


Tiling (Timothy Moy)

Tiling ended up being a no go as we didn't even have a use of implementing the shared memory. Since we weren't using shared memory, and tiling improves performance via shared memory we opted not to try implementing it and try other methods instead.

Block Size (Timothy Moy)

The code modified was line 22

   const int ntpb = 16; // number of threads per block

The first quick method to try and improve it was to change the block size. Playing with the block size changed the kernel run times, but it wasn't apparent what exactly causes it. Most likely it is due to the 16*16 block configuration being able to not take up all the memory of the SM, but is still large enough that it gives us a boost in execution times. https://devtalk.nvidia.com/default/topic/1026825/how-to-choose-how-many-threads-blocks-to-have-/

Media:Assign3-ntpb.png

In the end, a block size of 16 by 16 proved to be best for run times.

Moving Repeated Calculations to the Host (Timothy Moy)

I then tried merging the sinf() and cosf() function calls into one via sincosf() so that the kernel made less function calls. That proved to be trim the run times a bit, but then I noticed that sin and cos never change since our angle never changes. Thus, this led to testing of the sin and cos functions to use the Host to calculate it and pass them in as parameters for the kernel. The result was a much more significant run time since our kernel is no longer calculating the same number in each thread.

Kernel Signature Changes:

   __global__ void rotateKernel(int* oldImage, int* newImage, int rows, int cols, float rads) {

vs

   __global__ void rotateKernel(int* oldImage, int* newImage, int rows, int cols, /*float rads*/ float sinRads, float cosRads) {

Kernel Code Changes

   float sinRads = sinf(rads);
   float cosRads = cosf(rads);
   
   //float sinRads, cosRads;
   //__sincosf(rads, &sinRads, &cosRads);

vs

   //float sinRads = sinf(rads);
   //float cosRads = cosf(rads);
   
   float sinRads, cosRads;
   __sincosf(rads, &sinRads, &cosRads);

vs

   //float sinRads = sinf(rads);
   //float cosRads = cosf(rads);
   
   //float sinRads, cosRads;
   //__sincosf(rads, &sinRads, &cosRads);

and

Host Function Additions:

   float cos1 = cos(rads);
   float sin1 = sin(rads);

Kernel Launch Changed

   rotateKernel<<<dGrid, dBlock >>>(d_a, d_b, rows, cols, rads);

vs

   rotateKernel<<<dGrid, dBlock >>>(d_a, d_b, rows, cols, sin1, cos1);

The graph below shows the pronounced difference between the different sin cos methods.

 

There may be other variables that could be moved outside the kernel like r0 and c0, but due to time limitations they weren't tested.

All assignments compared in this file under "a1a2a3comps" sheet.

File:Assignment3 profile.xlsx.txt


Github

https://github.com/dleung25/GPU610-Assignment3-Image-Profiling

[We used the images in the Github and 100 degrees for all our tests]