Difference between revisions of "HeadCrab"

From CDOT Wiki
Jump to: navigation, search
(Created page with '{{GPU621/DPS921 Index | 20161}} = Intel Math Kernel Library (MKL) = == Team Member == # [mailto:rlanderson@senecacollege.ca?subject=dps921 Rene Anderson] == Intro == Intel Mat…')
 
(Intro)
 
(18 intermediate revisions by the same user not shown)
Line 3: Line 3:
 
== Team Member ==  
 
== Team Member ==  
 
# [mailto:rlanderson@senecacollege.ca?subject=dps921 Rene Anderson]
 
# [mailto:rlanderson@senecacollege.ca?subject=dps921 Rene Anderson]
 +
  
 
== Intro ==
 
== Intro ==
Intel Math Kernel Library [MKL]
+
 
 +
For this assignment I will be focusing on the Linear Algebra routines. I will use workshop six to demonstrate how BLAS can be used to significantly speed
 +
up the calculations and compare them to other parallelization methods.
 +
 
 +
<u>Compared to</u>
 +
#Serial
 +
#Cilk
 +
#Cilk with array notation and reduction
 +
#Cilk with SIMD and reduction
 +
#MKL CBLAS level 3
 +
#CUDA CUBLAS level 3
 +
#TBB
 +
#OpenMP
 +
 
 +
 
 +
'''Intel Math Kernel Library'''
 +
 
 +
MKL provides highly vectorized and threaded Linear Algebra, Fast Fourier Transforms,
 +
Vector Math and Statistics functions. Intel MKL uses industry standard APIs.
 +
This means that developers would have to make minor changes to their programs when
 +
switching to MKL.
 +
 
 +
 
 +
'''Intel MKL gives the developer control over the necessary trade-offs'''
 +
 
 +
#Result consistency vs performance
 +
#Accuracy vs performance
 +
 
 +
Intel MKL is also compatible with your choice of compilers, languages, operating systems, linking and threading models. One library solution across multiple environments means only one library to learn and manage.
 +
 
 +
 
 +
'''Linear Algebra'''
 +
 
 +
Intel MKL provides highly optimized BLAS routines
 +
#BLAS Level 1 vector-vector
 +
#BLAS Level 2 matrix-vector
 +
#BLAS Level 3 matrix-matrix
 +
 
 +
 
 +
'''How to enable Intel MKL'''
 +
 
 +
<u>#include <mkl.h></u>
 +
 
 +
<u>Command line</u>
 +
#-mkl           
 +
#-mkl=parallel to link with standard threaded Intel MKL.
 +
#-mkl=sequential to link with sequential version of Intel MKL.
 +
#-mkl=cluster to link with Intel MKL cluster components (sequential) that use Intel MPI.
 +
 
 +
<u>Microsoft Visual Studio</u>
 +
 
 +
project properties->Intel Performance Libraries->Intel Math Kernel Library
 +
 
 +
[[File:mkl.png]]
 +
 
 +
'''Fig 1 - Enable MKL'''
  
 
== Source code ==
 
== Source code ==
  
 +
[[File:mklcode.png]]
 +
 +
'''Fig 2 - MKL'''
 +
 +
 +
 +
 +
[[File:TBB.png]]
 +
 +
'''Fig 3 - TBB'''
 +
 +
 +
 +
 +
[[File:cuda.png]]
 +
 +
'''Fig 4 - CUDA'''
 +
 +
 +
 +
 +
[[File:omp.png]]
 +
 +
'''Fig 5 - OpenMP'''
  
 
== Useful Link ==
 
== Useful Link ==
 
# https://software.intel.com/en-us/intel-mkl/details
 
# https://software.intel.com/en-us/intel-mkl/details
 +
# https://software.intel.com/en-us/node/468380
 +
# https://software.intel.com/sites/default/files/managed/4a/d6/mkl_11.2.1_lnx_userguide.pdf
  
 
== Progress ==
 
== Progress ==
 +
 +
[[File:Data.png]]
 +
 +
'''Fig 6 - Recorded times'''
 +
 +
 +
 +
 +
[[File:rlachart.png]]
 +
 +
'''Fig 7 - Graph of times'''

Latest revision as of 11:31, 5 April 2016


GPU621/DPS921 | Participants | Groups and Projects | Resources | Glossary

Intel Math Kernel Library (MKL)

Team Member

  1. Rene Anderson


Intro

For this assignment I will be focusing on the Linear Algebra routines. I will use workshop six to demonstrate how BLAS can be used to significantly speed up the calculations and compare them to other parallelization methods.

Compared to

  1. Serial
  2. Cilk
  3. Cilk with array notation and reduction
  4. Cilk with SIMD and reduction
  5. MKL CBLAS level 3
  6. CUDA CUBLAS level 3
  7. TBB
  8. OpenMP


Intel Math Kernel Library

MKL provides highly vectorized and threaded Linear Algebra, Fast Fourier Transforms, Vector Math and Statistics functions. Intel MKL uses industry standard APIs. This means that developers would have to make minor changes to their programs when switching to MKL.


Intel MKL gives the developer control over the necessary trade-offs

  1. Result consistency vs performance
  2. Accuracy vs performance

Intel MKL is also compatible with your choice of compilers, languages, operating systems, linking and threading models. One library solution across multiple environments means only one library to learn and manage.


Linear Algebra

Intel MKL provides highly optimized BLAS routines

  1. BLAS Level 1 vector-vector
  2. BLAS Level 2 matrix-vector
  3. BLAS Level 3 matrix-matrix


How to enable Intel MKL

#include <mkl.h>

Command line

  1. -mkl
  2. -mkl=parallel to link with standard threaded Intel MKL.
  3. -mkl=sequential to link with sequential version of Intel MKL.
  4. -mkl=cluster to link with Intel MKL cluster components (sequential) that use Intel MPI.

Microsoft Visual Studio

project properties->Intel Performance Libraries->Intel Math Kernel Library

Mkl.png

Fig 1 - Enable MKL

Source code

Mklcode.png

Fig 2 - MKL



TBB.png

Fig 3 - TBB



Cuda.png

Fig 4 - CUDA



Omp.png

Fig 5 - OpenMP

Useful Link

  1. https://software.intel.com/en-us/intel-mkl/details
  2. https://software.intel.com/en-us/node/468380
  3. https://software.intel.com/sites/default/files/managed/4a/d6/mkl_11.2.1_lnx_userguide.pdf

Progress

Data.png

Fig 6 - Recorded times



Rlachart.png

Fig 7 - Graph of times