Fastest Fourier Transform in the West to make all type of Fourier transforms.

The 3.3.3 version is installed in /software/fftw. Several types have been compiled (threads, simple precission,etc)

To link them use

-L/software/fftw -lfftw3

If you have doubts ask the technicians.

More information in the web page FFTW.

General information

Based on C and FORTRAN directives, HMPP Workbench version 2.5.2 offers a high level abstraction for  GPGPUs. HMPP compiler integrates powerful data-parallel backends for NVIDIA CUDA and OpenCL. The HMPP runtime ensures application deployment on multi-GPU systems.

How to use

In order to execute the compiler use

hmpp

for example, to compile the test.c program by using the gcc compiler use

hmpp gcc test.c -o test

More information

HMPP web page

Compilers and installed versions in the different architectures.

Compilers	Xeon	Itanium	Opteron	Pendulo
GNU
[intlink id=”454″ type=”post”]Intel[/intlink]
[intlink id=”3664″ type=”post”]Portlan (PGI)[/intlink]	10.9		10.9
[intlink id=”3672″ type=”post”]Cuda[/intlink]	3.2
Java	Java HotSpot(TM) Server VM (build 1.6.0_03-b05, mixed mode)  OpenJDK 64-Bit Server VM (build 14.0-b16, mixed mode)	Java JRockit(R) (build R27.6.3-40)	Java Java HotSpot(TM) 64-Bit Server VM  (build 14.0-b16, mixed mode)
[intlink id=”5454″ type=”post”]HMPP[/intlink]
MPI	Xeon	Itanium	Opteron	Pendulo
[intlink id=”459″ type=”post”]HP-MPI[/intlink]
[intlink id=”459″ type=”post”]Bull-MPI[/intlink]
[intlink id=”459″ type=”post”]Intel MPI[/intlink]
[intlink id=”459″ type=”post”]Openmpi [/intlink]
[intlink id=”459″ type=”post”]mvapich2 1.4.1 [/intlink]
Libraries	Xeon	Itanium	Opteron	Pendulo
[intlink id=”447″ type=”post”]Intel MKL[/intlink]
[intlink id=”462″ type=”post”] FFTW[/intlink]
[intlink id=”3775″ type=”post”] GSL[/intlink]