Hey Philippe,
I've disabled LU and renamed sizes to {M,N,K}. It's running fine now.
Except custom sparse matrices :D I have no idea how to load and use them
for benchmarking, since the new procedure using on-the-fly matrix
generation is very different than the one using boost. Here's how I tried
to it:
std::vector< std::map<unsigned int, ScalarType> > stl_A;
viennacl::tools::sparse_matrix_adapter<ScalarType> adapted_A(stl_A);
if(customSparseMatrixPath.isEmpty()){
//no custom matrix specified, generate the default one
viennacl::tools::generate_fdm_laplace(adapted_A, xPoints, yPoints );
}
else{
//load matrix from file
if(!viennacl::io::read_matrix_market_file(adapted_A /* Is this the
right variable to read into? */, customSparseMatrixPath.toStdString()) ){
std::cout << "Custom sparse matrix read successfully" <<std::endl;
}
else{
std::cout << "Failed to read custom sparse matrix" <<std::endl;
}
}
It always gives either unsupported format or plain failed to read matrix.
Some help would be greatly appreciated.
By the way, I'm starting to feel minor addiction when playing with the
expert mode. There's certainly something very attracting about tweaking
those numbers to get bigger numbers :D Except in the case of blas3. I just
couldn't find a better matrix size combination than the default
{1920,1920,1920}. Are these sizes the result of that magic auto-tuner
thingy, or is it just unlucky tweaking? :D
Regards, Namik
On Mon, Aug 18, 2014 at 4:59 AM, Philippe Tillet <phil.til...@gmail.com>
wrote:
> Hey Namik,
>
> The code looks fine. As a small tip, I would advise to use
> blas3MatrixSize{A,B,C} = {M, N, K} ; it's much more conventional. I would
> also suggest to remove LU from the benchmark. I only achieve 11 GFLOP/s on
> my machine (GEMM peaks at 120GFLOP/s). It will smash the overall score if
> you keep it enabled!
>
> Philippe (Not sleeping either :-p)
>
>
> 2014-08-17 23:28 GMT+02:00 Namik Karovic <namik.karo...@gmail.com>:
>
>> Hi all,
>>
>> I just pushed the first working version of expert(custom) benchmark mode.
>> Selecting custom sparse matrices is yet to be implemented, but all other
>> benchmark configs are working.
>>
>> Except blas3, that is. I think I got the sizes wrong. I'd appreciate it
>> if someone could check if I did it right:
>>
>> //blas3MatrixSizeA,B = size1,2
>> //blas3MatrixSizeB,C = size2,3
>> viennacl::matrix<ScalarType> vcl_A(blas3MatrixSizeA, blas3MatrixSizeB);
>> viennacl::matrix<ScalarType> vcl_B(blas3MatrixSizeB, blas3MatrixSizeC);
>> viennacl::matrix<ScalarType> vcl_C(blas3MatrixSizeA, blas3MatrixSizeC);
>>
>> // Fill the matrix
>> for (unsigned int i = 0; i < blas3MatrixSizeA; ++i)
>> for (unsigned int j = 0; j < blas3MatrixSizeB; ++j)
>> stl_A[i*blas3MatrixSizeA + j] = random<ScalarType>();
>>
>> for (unsigned int i = 0; i < blas3MatrixSizeB; ++i)
>> for (unsigned int j = 0; j < blas3MatrixSizeC; ++j)
>> stl_B[i + j*blas3MatrixSizeC] = random<ScalarType>();
>>
>> //using ranges
>> viennacl::range r(blas3MatrixSizeB/4, 3 * blas3MatrixSizeB/4);
>>
>> //using slices
>> viennacl::slice s(0, 2, blas3MatrixSizeB/2);
>>
>> The benchmark crashes on test 4 (LU factorization). I don't know if I
>> messed up somewhere before test 4 (in the code written above), or somewhere
>> else.
>>
>> Regards, Namik
>>
>>
>>
>> ------------------------------------------------------------------------------
>>
>> _______________________________________________
>> ViennaCL-devel mailing list
>> ViennaCL-devel@lists.sourceforge.net
>> https://lists.sourceforge.net/lists/listinfo/viennacl-devel
>>
>>
>
------------------------------------------------------------------------------
_______________________________________________
ViennaCL-devel mailing list
ViennaCL-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/viennacl-devel
