Search found 913 matches

by mgates3
Tue May 15, 2012 3:49 pm
Forum: User discussion
Topic: MAGMA 1.2 Released
Replies: 7
Views: 14917

Re: MAGMA 1.2 Released

Yes, a few issues with the MAGMA 1.2 release have been corrected. I will post a new tar file hopefully later today. The [sdcz]hetrd_mgpu.cpp and [sdcz]latrd_mgpu.cpp files were not intended to be released yet. They should have been removed from the Makefile as well. The [sdc] versions of wrappers we...
by mgates3
Tue May 15, 2012 2:18 pm
Forum: User discussion
Topic: clMAGMA problems with CUDA
Replies: 4
Views: 2305

Re: clMAGMA problems with CUDA

clMAGMA 0.1 is for OpenCL. The regular MAGMA release is for CUDA (i.e., MAGMA 1.1). The #defines you see in clMAGMA are essentially to make existing CUDA code compile with OpenCL.

-mark
by mgates3
Sat May 05, 2012 1:32 pm
Forum: User discussion
Topic: Problem with testing_zgesv
Replies: 12
Views: 9420

Re: Problem with testing_zgesv

Nothing is apparent yet. Since dgetrf works, it doesn't seem to be a precision issue. Since smaller sizes work, it doesn't seem to be a memory issue. It looks like everything double complex is failing for you. Try testing_*gemm with various sizes. What is your hardware (CPU, memory, GPU), OS, compil...
by mgates3
Fri May 04, 2012 6:35 pm
Forum: User discussion
Topic: CUDA and CUBLAS initialization using MAGMA
Replies: 2
Views: 3864

Re: CUDA and CUBLAS initialization using MAGMA

Yes, we currently recommend initializing as in the testings.h file. At the moment, a number of MAGMA functions call the CUDA driver API (functions starting with cu instead of cuda). cuInit() and cuCtxCreate() must be called before calling the CUDA driver API. In a future release, MAGMA's use of thes...
by mgates3
Fri May 04, 2012 5:58 pm
Forum: User discussion
Topic: Problem with testing_zgesv
Replies: 12
Views: 9420

Re: Problem with testing_zgesv

Can you try other sizes, such as
testing_zgetrf -M 1000 -N 1000
testing_zgetrf -M 2000 -N 2000
testing_zgetrf -M 4000 -N 4000
Do other double complex routines work, such zpotrf, zgeqrf, etc.?
-mark
by mgates3
Fri May 04, 2012 5:51 pm
Forum: User discussion
Topic: While compiling clMAGMA getting an error:
Replies: 1
Views: 1161

Re: While compiling clMAGMA getting an error:

Did you add the AMD BLAS directory to your library path? E.g., in csh,

setenv LD_LIBRARY_PATH ${LD_LIBRARY_PATH}:/opt/clAmdBlas-1.7.245/lib64

-mark
by mgates3
Fri May 04, 2012 5:46 pm
Forum: User discussion
Topic: magma_dgeev
Replies: 2
Views: 1880

Re: magma_dgeev

The magma dgeev code is a hybrid algorithm -- it uses both the CPU and the GPU to solve the system. Currently magma provides only the interface that takes the data on the CPU. Internally it copies blocks to the GPU during certain operations such as the Hessenberg factorization, while other operation...
by mgates3
Fri May 04, 2012 3:43 pm
Forum: User discussion
Topic: SGEMM when beta=0
Replies: 5
Views: 3714

Re: SGEMM when beta=0

Here are some results on a Tesla T20 (Fermi), Magma 1.1, CUDA 4.0, Intel MKL. (Source code attached.) The first column of C is set to NAN on input. When beta=0, implementations differ on whether they propagate NAN values or not. In particular, MAGMA BLAS does propagate NAN values. Therefore, you do ...
by mgates3
Fri May 04, 2012 10:49 am
Forum: User discussion
Topic: SGEMM when beta=0
Replies: 5
Views: 3714

Re: SGEMM when beta=0

When you didn't initialize the matrix, what result were you getting, and what result did you expect? A short sample code would be helpful. Technically, the beta*C must be carried out to properly propagate NAN values that may be in the C matrix. I'll look into how some different libraries handle this...
by mgates3
Fri May 04, 2012 10:43 am
Forum: User discussion
Topic: Using sgemm for rectangular (non-square) matrix multiply
Replies: 7
Views: 8828

Re: Using sgemm for rectangular (non-square) matrix multiply

The gemm in CUBLAS (and in MAGMA BLAS) works for rectangular matrices. As far as I know, all the CUBLAS work for all matrix sizes. If you are seeing some problems, please report the specific case. A short sample code would be helpful.
-mark