Archives Discussions

[twinclouds]

Do you mean 9.2 because I think 9.2 is not available yet?

[twinclouds]

Sorry, I meant 9.3 is not available yet.

[rahulgarg]

As someone suggested earlier, you dont need a full installation of gfortran 4.1. Instead search for a deb of libgfortran1, extract it into a folder and then include that folder in your library path.

[twinclouds]

Thanks.  Will try.

[twinclouds]

Not sure where to ask and don't know if anyone have asked this question before.  Can program using ACML-GPU run in virtual machine under VMWARE server?

[twinclouds]

Dear Michael:

I eventually made the acml-GPU examples work on my Linux machine.  What a relief!

Acutally, some of the programs in the example folder are in infortran, so gfortran-4.1.2 is needed if one want to comple them.   I downgraded my Ubuntu to 8.04 for whcih the gcc4.1.2 and gfortran-4.1.2 are both available and everything works.

What I really want to use are the FFT routines in the acml library.  It looks like that they are not available for acml-GPU library yet.  If it is true, do you know when they will be available?  If possible I wish I can get some early vesions of them to test out.  I will be happy to report the results back once I try them out.  Mainly, I am interested in 1D FFT routines.

Thanks in advance for your help.

[indi123]

Only 64 bit versions are posted. Will there be a 32 bit version?

[tomidomi]

I have the same question like twinclouds. What about FFT? I am a physicist and most often used procedures by me are 3D FFT and eigenvalue problems (LAPACK solvers). From scientific point of view I need double precision usually complex floating point procedures. Double precision because my simulations take millions time steps and single precision is not acceptable.

[twinclouds]

I recently wrote a program to benchmark NVidia's CUDA FFT (CUFFT) routine.  The program did 204800 2048 point FFT and IFFT pair.  It take about 17 seconds using a 9600GT.  This is similar to the routines in FFTw Library on a 2.66 GHz Core 2 CPU.  Very disappointed.  Actually, AMCL scored the best.  It only took 7 seconds.  I also found independent verification from the Web that CUFFT only became faster if the FFT size is larger than 32768 points.

Hope AMCL-GPU FFT will be better when ready.