Chance to use modern Graphics Cards as.. - Page 3

Turkmenistan · 2003-05-01, 15:56

I found interesting software experimentation with GPU support.
http://www.sourceforge.net/projects/ffff

Here is the part off the readme.

TKM

=========================================================
FFFF - Fast Floating Fractal Fun v3.2.0
Author: Daniele Paccaloni (daniele.paccaloni@dylogic.com)
Project page: http://www.sourceforge.net/projects/ffff
=========================================================

FFFF is just the fastest mandelbrot generator for Win32.
It still uses brute force, every pixel you see has been calculated. Future versions will support guessing.
And it is GNU GPL OpenSource.

HISTORY
-------
Authors between square brackets.

v3.2.0

- Added GPU OpenGL ARB Fragment Program calculation (couldn't yet get less than 3 instructions per iter on the ATI R300) [DP].
- Added a decent mouse center zoom feature. [DP]
- Benchmarking fixes [DP].
- Minor fixes [DP].

v3.1.2

- Optimized SSE calculation code [GB].
- Added SSE2 support (DualPixel, double precision) [DP].
- Calculation mode keys are changed (see help text), press '0' for GPU calc [DP].
- Minor fixes [DP].

v3.1.1

- Fixed a *stupid* bug that caused crashing of benchmarks on Intel CPUs (sorry !!) [DP].
- Fixed a *stupid* bug that caused crashing of benchmarks on systems without a GPU (sorry again !!) [DP].
- Documented the new orbit draw mode (press o and move the mouse in single buffer mode) [DP].

v3.1.0

- Added AMD 3DNow! dual pixel calculation [GB].
- Added Radeon GPU pixel calculation using GL_VERTEX_SHADER_EXT (see notes). Maybe it works on other GPUs, tell me ! [DP].
- Added experimental GPU benchmarking (does not work on my Radeon 9000..) [DP].
- Slightly improved speed of SSE2 quad pixel calculation by constant data alignment on 64 bytes boundaries [DP].
- Slightly improved speed FPU code.
- Code restructured to resemble an object oriented program (quite not yet :).
- Fixed minor bugs [DP, thanks for your bug reports].

...

Dresdenboy · 2003-05-05, 11:00

Another interesting paper:

"Fast Matrix Multiplies Using Graphics Hardware"
http://www.mersenneforum.org/attachm...pap.pap313.pdf
(I found this link on http://yosirin9.hp.infoseek.co.jp/)

Regards,
Dresdenboy
(M.Waldhauer)

willmore · 2003-05-13, 15:27

With every major new generation of graphics hardware we see this question asked again. Some just dismiss it as it *has* been asked so many times. But, this time, things have changed. GPUs are more general purpose in archetecture. They are becoming progammable--in the tranditional sense of the word. They're even supporting floating point data types. I think we'll see some use from them in coming years for real mathetical work. Maybe not LL tests, but trial factoring, or such, surely.

To clarify some things. They do not support 128 bit FP operations. :) They support a more SSE kind of 4x32bit FP operations. The thing that does make them look good is that they're configured for a lot of trig transforms and matrix math. That might be useful for some applications.

In the future, I think we'll see graphics units getting more and more generic and more and more useful for tasks like this.

This could be fun. :)

clowns789 · 2003-06-16, 20:10

How do you change the speed of Prime95? That's all I use that computer for now and it's rather slow ( about .6 seconds/ iteration). At least it's consistent unlike a lot of other people.

Clowns789, your "Prime Minister"

Dresdenboy · 2003-06-17, 06:50

It's speed is dependend on free CPU time. Usually it runs as fast as possible without disturbing other apps.

But the time per iteration depends on speed of CPU (not raw GHz, but how many floating point or SSE2 calculations it can do per second - all CPU families are different in this manner) and on the task to do. Testing a smaller number is faster than a big one.

DDB

only_human · 2003-10-31, 09:52

I've come across a site devoted to cataloging the use of Graphics Processing Units for general purpose computation. There is a link to a forum too:
www.gpgpu.org
General Purpose Computation on GPUs

Quote:

With the increasing programmability of commodity graphics processing units (GPUs), these chips are capable of performing more than the specific graphics computations for which they were designed. They are now capable coprocessors, and their high speed makes them useful for a variety of applications. The goal of this page is to catalog the current and historical use of GPUs for general purpose computation.

QuintLeo · 2003-11-02, 23:21

I find it interesting that one application on that GPU page specifically mentions FFT work....

8-)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Prime95 and graphics cards	keithschmidt	Information & Answers	45	2016-09-10 10:08
Modern C	Dubslow	Programming	15	2016-01-12 09:13
New Linux rootkit leverages graphics cards for stealth.	swl551	Lounge	0	2015-05-08 14:06
Nvidia's next-generation graphics cards	ixfd64	GPU Computing	22	2014-11-15 04:25
how do graphics cards work so fast?	ixfd64	Hardware	1	2004-06-02 03:01

2003-05-01, 15:56	#23
Turkmenistan May 2003 1₁₀ Posts	I found interesting software experimentation with GPU support. http://www.sourceforge.net/projects/ffff Here is the part off the readme. TKM ========================================================= FFFF - Fast Floating Fractal Fun v3.2.0 Author: Daniele Paccaloni (daniele.paccaloni@dylogic.com) Project page: http://www.sourceforge.net/projects/ffff ========================================================= FFFF is just the fastest mandelbrot generator for Win32. It still uses brute force, every pixel you see has been calculated. Future versions will support guessing. And it is GNU GPL OpenSource. HISTORY ------- Authors between square brackets. v3.2.0 - Added GPU OpenGL ARB Fragment Program calculation (couldn't yet get less than 3 instructions per iter on the ATI R300) [DP]. - Added a decent mouse center zoom feature. [DP] - Benchmarking fixes [DP]. - Minor fixes [DP]. v3.1.2 - Optimized SSE calculation code [GB]. - Added SSE2 support (DualPixel, double precision) [DP]. - Calculation mode keys are changed (see help text), press '0' for GPU calc [DP]. - Minor fixes [DP]. v3.1.1 - Fixed a stupid bug that caused crashing of benchmarks on Intel CPUs (sorry !!) [DP]. - Fixed a stupid bug that caused crashing of benchmarks on systems without a GPU (sorry again !!) [DP]. - Documented the new orbit draw mode (press o and move the mouse in single buffer mode) [DP]. v3.1.0 - Added AMD 3DNow! dual pixel calculation [GB]. - Added Radeon GPU pixel calculation using GL_VERTEX_SHADER_EXT (see notes). Maybe it works on other GPUs, tell me ! [DP]. - Added experimental GPU benchmarking (does not work on my Radeon 9000..) [DP]. - Slightly improved speed of SSE2 quad pixel calculation by constant data alignment on 64 bytes boundaries [DP]. - Slightly improved speed FPU code. - Code restructured to resemble an object oriented program (quite not yet :). - Fixed minor bugs [DP, thanks for your bug reports]. ...

2003-05-05, 11:00	#24
Dresdenboy Apr 2003 Berlin, Germany 19² Posts	Another interesting paper: "Fast Matrix Multiplies Using Graphics Hardware" http://www.mersenneforum.org/attachm...pap.pap313.pdf (I found this link on http://yosirin9.hp.infoseek.co.jp/) Regards, Dresdenboy (M.Waldhauer)

2003-05-13, 15:27	#25
willmore Aug 2002 61 Posts	With every major new generation of graphics hardware we see this question asked again. Some just dismiss it as it has been asked so many times. But, this time, things have changed. GPUs are more general purpose in archetecture. They are becoming progammable--in the tranditional sense of the word. They're even supporting floating point data types. I think we'll see some use from them in coming years for real mathetical work. Maybe not LL tests, but trial factoring, or such, surely. To clarify some things. They do not support 128 bit FP operations. :) They support a more SSE kind of 4x32bit FP operations. The thing that does make them look good is that they're configured for a lot of trig transforms and matrix math. That might be useful for some applications. In the future, I think we'll see graphics units getting more and more generic and more and more useful for tasks like this. This could be fun. :)

2003-06-16, 20:10	#26
clowns789 Jun 2003 The Computer 401 Posts	How do you change the speed of Prime95? That's all I use that computer for now and it's rather slow ( about .6 seconds/ iteration). At least it's consistent unlike a lot of other people. Clowns789, your "Prime Minister"

2003-06-17, 06:50	#27
Dresdenboy Apr 2003 Berlin, Germany 19² Posts	It's speed is dependend on free CPU time. Usually it runs as fast as possible without disturbing other apps. But the time per iteration depends on speed of CPU (not raw GHz, but how many floating point or SSE2 calculations it can do per second - all CPU families are different in this manner) and on the task to do. Testing a smaller number is faster than a big one. DDB

2003-11-02, 23:21	#29
QuintLeo Oct 2002 Lost in the hills of Iowa 1C0₁₆ Posts	I find it interesting that one application on that GPU page specifically mentions FFT work.... 8-)