Part of my Ph.D. research involves solving coupled reaction diffusion equations on arbitrary closed surfaces using Finite Element Method. In the past I had been using the ?GESV subroutine to solve the resulting systems of equations, but due to the large resolution of my meshes, this became impractical to use, especially since the matrix is sparse. I then discovered Intel's sparse solvers and have been attempting to implement the DSS routine.
I have a problem using the data fitting for a linear interpolation of a vector valued function. It seems that the format rhint is ignored by the function dfdInterpolate1D.
I have encountered a confirmed bug with routine mkl_zcsrcoo in 11.0.4. It is my understanding that this bug has been fixed in 11.0.5. Would be it possible for me to get 11.0.5 ?
The MKL version that I have has bugs, and I need to update it. I need help getting it updated.
I am confused about what version of Intel Fortran I have installed.
The Help About box in Visual Studio says: Intel(R) Visual Fortran Composer XE 2013 Update 4 Integration for Microsoft Visual Studio* 2012, 13.0.3624.11
When compiling code, the output window says: Compiling with Intel(R) Visual Fortran Compiler XE 22.214.171.124
The MKL library that I have calls itself: Intel(R) Math Kernel Library Version 11.0.4 Product Build 20130517
I am finding that pardiso does not work for a 1 element matrix. Calling pardiso with phase=11 returns error=0, which means success, but the pt(:) pointer array is all zero's. So when I subsequently call pardiso with phase=23, it fails with error=-7.
Can someone please confirm that pardiso does or does not work for a 1 element matrix?
Intel® Parallel Studio XE 2015 Update 5 Professional Edition for Fortran and C++ parallel software development suite combines Intel's C/C++ compiler and Fortran compiler; performance and parallel libraries; error checking, code robustness, and performance profiling tools into a single suite offering. This new product release includes:
I am trying to solve a system with `cluster_sparse_solver`, and am getting unexpected results. https://gist.github.com/ivan-krukov/157372d9a55db244c4b4
In my case, given a sparse matrix `A`, we want to solve `Ax=e`, where `e` is the first column of the identity matrix with the appropriate size (A column where the first entry is 1, rest are zeros).
I am doing development on a 24-core machine (E5-2697-v2). When I launch a single DGEMM where the matrices are large (m=n=k=15,000), the performance improves as I increase the number of threads used, which is expected. For reference, I get about 467 GFLOPs/sec using 24 cores.