readme.txt

基于GPU进行快速科学计算

GPUFFTW v1.0

/******************************************************************************\

  Copyright 2005 The University of North Carolina at Chapel Hill.
  All Rights Reserved.

  Permission to use, copy, modify and distribute this software and its
  documentation for educational, research and non-profit purposes, without
  fee, and without a written agreement is hereby granted, provided that the
  above copyright notice and the following three paragraphs appear in all
  copies. Any use in a commercial organization requires a separate license.

  IN NO EVENT SHALL THE UNIVERSITY OF NORTH CAROLINA AT CHAPEL HILL BE LIABLE
  TO ANY PARTY FOR DIRECT, INDIRECT, SPECIAL, INCIDENTAL, OR CONSEQUENTIAL
  DAMAGES, INCLUDING LOST PROFITS, ARISING OUT OF THE USE OF THIS SOFTWARE AND
  ITS DOCUMENTATION, EVEN IF THE UNIVERSITY OF NORTH CAROLINA HAVE BEEN
  ADVISED OF THE POSSIBILITY OF SUCH DAMAGES.


  Permission to use, copy, modify and distribute this software and its
  documentation for educational, research and non-profit purposes, without
  fee, and without a written agreement is hereby granted, provided that the
  above copyright notice and the following three paragraphs appear in all
  copies.

  THE UNIVERSITY OF NORTH CAROLINA SPECIFICALLY DISCLAIM ANY WARRANTIES,
  INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND
  FITNESS FOR A PARTICULAR PURPOSE.  THE SOFTWARE PROVIDED HEREUNDER IS ON AN
  "AS IS" BASIS, AND THE UNIVERSITY OF NORTH CAROLINA HAS NO OBLIGATION TO
  PROVIDE MAINTENANCE, SUPPORT, UPDATES, ENHANCEMENTS, OR MODIFICATIONS.


   ---------------------------------
  |Please send all BUG REPORTS to:  |
  |                                 |
  |   [e-mail]               |
  |                                 |
   ---------------------------------


  The authors may be contacted via:

  US Mail:         N. Govindaraju,
                       Department of Computer Science
                       Sitterson Hall, CB #3175
                       University of North Carolina
                       Chapel Hill, NC 27599-3175

\*****************************************************************************/


------------
DIRECTORIES:
------------
src: source for GPUFFTW library
include : include files for applications using GPUFFTW
Samples: example demonstrating the usage of our library
	- Example3 - FFT performance on complex and real 1-D arrays using IEEE 32-bit floating point values
Project files: Project files for Microsoft Visual Studio 6.0, Microsoft Visual Studio .NET 2003 and higher, and makefiles for linux
lib: Contains the directories for libraries for WIN32 and linux

-----------------
API DOCUMENTATION: Refer to http://gamma.cs.unc.edu/GPUFFTW
-----------------

--------------------
SYSTEM REQUIREMENTS:
--------------------
1. OS: Microsoft Windows XP/2000 and Linux
2. RAM: Atleast the size of graphics processor video memory is required.
3. GPU: NVIDIA GeForce/Quadro family card with support for the following OpenGL extensions:
	ARB_texture_rectangle
        ARB_fragment_program
	EXT_framebuffer_object

	The above requirements are met by NV30-based GPUs and above (GeForce FX and GeForce 6 series)
	The library has been tested on the following cards:
    
      a. GeForce 7900/7800/6800 Ultra/GTO
      b. GeForce 5950 Ultra
      c. GeForce 5800
      d. Quadro FX 4000 
      
      Laptop graphics cards:
      e. QuadroFX 700 Go, QuadroFX 1000 Go, QuadroFX 1400 Go, GeForce 6800 Go, GeForce 7800Go

      For obtaining reasonably high performance, we recommend a PC with AGP8X/PCI-Express NVIDIA GeForce 7800 GT or faster GPUs.
	
4. Video RAM: The Video RAM will determine the maximum array length that can be used to perform FFT on the GPU. A rough guideline
for sorting 32-bit floats is: Maximum array length in millions = Video RAM in MB / 32 for real values and video RAM/16 for complex values. 
Therefore, on a card with 256 MB VRAM, the maximum-length array which can be sorted is 256/32 = 8 Million values


5. Drivers: Latest drivers from NVIDIA (version 7772 or higher for windows, and 7664 for linux).

Note:
  1. ATI cards: ATI cards are not supported in the present release of GPUFFTW mainly due to the lack of suport for ARB_texture_rectangle in fragment programs on
 current ATI drivers. These cards will be supported in future releases. 
  
  2. Non power-of-two arrays: The present implementation will not perform FFT on arrays with non power-of-two size.

  3. Higher dimensional FFTs: Our current implementation does not support higher dimensional FFTs. There is sample code with missing code fragments for 2D FFTs if you prefer experimenting. 

------------
COMPILATION:
------------

VC 6.0: The workspace contains the library project and the two examples. Make sure to build the library before the examples.
VS .NET: Rebuild the solution and run the examples by choosing the project for an example as the "startup project".
Linux: Compile gpufftw by using the command "make". Modify the makefile to include the library path containing libglut.a and its include files "glut.h". The examples can be built using the command "make example3".