Difference between revisions of "FFTW"

From UFRC
Jump to navigation Jump to search
(18 intermediate revisions by 7 users not shown)
Line 1: Line 1:
==Description==
+
[[Category:Software]][[Category:Numerical]]
From the [http://www.fftw.org/index.html FFTW website]:
+
{|<!--CONFIGURATION: REQUIRED-->
<blockquote>
+
|{{#vardefine:app|FFTW}}
FFTW is a C subroutine library for computing the discrete Fourier transform (DFT) in one or more dimensions, of arbitrary input size, and of both real and complex data (as well as of even/odd data, i.e. the discrete cosine/sine transforms or DCT/DST). We believe that FFTW, which is free software, should become the FFT library of choice for most applications.
+
|{{#vardefine:url|http://http://www.fftw.org}}
</blockquote>
+
<!--CONFIGURATION: OPTIONAL (1=ON)-->
 +
|{{#vardefine:exe|}} <!--RUNNING -->
 +
|{{#vardefine:conf|1}} <!--CONFIGS-->
 +
|{{#vardefine:pbs|}} <!--PBS SCRIPTS-->
 +
|{{#vardefine:policy|}} <!--POLICY-->
 +
|{{#vardefine:testing|}} <!--TESTING/PROFILING-->
 +
|{{#vardefine:faq|1}} <!--FAQ-->
 +
|{{#vardefine:citation|1}} <!--CITATION-->
 +
|{{#vardefine:installation|}} <!--INSTALLATION-->
 +
|}
 +
<!--BODY-->
 +
<!--Description-->
 +
{{#if: {{#var: url}}|
 +
{{App_Description|app={{#var:app}}|url={{#var:url}}|name={{#var:app}}}}|}}
 +
Fastest Fourier Transform in the West, is a C library for computing the Discrete Fourier Transform (DFT) in one or more dimensions, of arbitrary input size, and of both real and complex data. Developed at MIT, it is by far the fastest implementation of FFT algorithm (upheld by regular benchmarks.)
  
==2.1.5==
+
<!--Modules-->
Compiled with: Intel compilers
+
==Required Modules==
 +
[[Modules|Modules documentation]]
 +
===Serial with Intel Compiler ===
 +
*intel/2016.0.109
 +
*fftw
  
Options: --enable-threads
+
===Serial with GNU Compiler ===
 +
*gcc/4.7.2
 +
*fftw
  
Location: /opt/intel/fftw/2.1.5/lib
+
===Parallel with Intel Compiler ===
 +
*intel/2016.0.109
 +
*openmpi
 +
*fftw
 +
<!--Additional-->
 +
{{#if: {{#var: exe}}|==Additional Information==
 +
DOUBLE_CLICK_TO_WRITE_INSTRUCTIONS_ON_RUNNING_THE_ACTUAL_BINARY
 +
|}}
 +
<!--Citation-->
 +
{{#if: {{#var: citation}}|==Citation==
 +
If you publish research that uses {{#var:app}} you may want to cite it as follows:<br/><br/>
 +
"[http://www.fftw.org/pldi99.pdf A Fast Fourier Transform Compiler]," by Matteo Frigo, in the Proceedings of the
 +
1999 ACM SIGPLAN Conference on Programming Language Design and Implementation
 +
(PLDI '99), Atlanta, Georgia, May 1999
 +
|}}
 +
 
 +
<!--Turn the Table of Contents and Edit paragraph links ON/OFF-->
 +
__NOTOC____NOEDITSECTION__

Revision as of 20:51, 5 February 2016

Description

FFTW website  
Fastest Fourier Transform in the West, is a C library for computing the Discrete Fourier Transform (DFT) in one or more dimensions, of arbitrary input size, and of both real and complex data. Developed at MIT, it is by far the fastest implementation of FFT algorithm (upheld by regular benchmarks.)

Required Modules

Modules documentation

Serial with Intel Compiler

  • intel/2016.0.109
  • fftw

Serial with GNU Compiler

  • gcc/4.7.2
  • fftw

Parallel with Intel Compiler

  • intel/2016.0.109
  • openmpi
  • fftw

Citation

If you publish research that uses FFTW you may want to cite it as follows:

"A Fast Fourier Transform Compiler," by Matteo Frigo, in the Proceedings of the 1999 ACM SIGPLAN Conference on Programming Language Design and Implementation (PLDI '99), Atlanta, Georgia, May 1999