Search Results for "cuda"

LightSpMV is a novel CUDA-compatible sparse matrix-vector multiplication (SpMv) algorithm using the standard compressed sparse row (CSR) storage format. We have evaluated LightSpMV using various sparse matrices and further compared it to the CSR-based SpMV subprograms in the state-of-the-art CUSP and cuSPARSE. Performance evaluation reveals that on a single Tesla K40c GPU, LightSpMV is superior to both CUSP and cuSPARSE, with a speedup of up to 2.60 and 2.63 over CUSP, and up to 1.93 and 1.79 over cuSPARSE for single and double precision, respectively.

MXLib is a C++ wrapper around the Intel® Integrated Performance Primitives (IPP) library and NVidia NPP CUDA library. You can use either IPP code (or a subset of functions that do not require IPP) on the CPU side, or use NPP/CUDA on the GPU side, or use both together. The function syntax is similar to that found in MatLab and the library is designed to make it easy to port your code from MatLab to C++. The idea is to provide Scientists, Engineers, Researchers and other non full-time programmers an easy to use, high performance library of functions.

...After selection of the initial frequency candidates, the algorithm passes through all their possible combinations and estimates their multi-frequency statistical significance. In the end, it prints out the set of largest frequency tuples that were still found significant. The GPU computing is implemented through CUDA and brings a significant performance increase. It is still possible to run FreDec solely on CPU, if no suitable GPU device is available in the system. See the details of the underlying theory in Baluev 2013, MNRAS, V. 436, P. 807 The description of the algorithm itself can be found in arXiv:1309.0100. Please go to the project forum to report any bugs.