Home > High Performance Computing > HPC Libraries

HPC Libraries

CRL is building High Performance Clusters of several hundred to thousands of nodes. Ideally, an all-connect network will provide complete connectivity. However, the number of connections increases quadratically O(n2) with the number of nodes (n). To solve this problem we have developed an unique architecture which has O(n) connections for n nodes. In the parallel libraries group we implement optimal parallel functions to support a variety of applications. These include - Linear Matrix Algebra functions to support applications in Computational Fluid Dynamics (CFD) and NanoScience Area. These include basic operations such as matrix-matrix and sparse matrix-vector multiplications, various sparse solvers, etc. - Fast Parallel Search to support applications in Life Sciences Area and Electronic CAD.