The Intel Adaptive Spike-Based Solver is a poly-algorithm that uses many different strategies to solve large banded systems in parallel. It uses a novel decomposition method to balance computation against communication requirements. Since the characteristics of the input matrix (e.g., size, bandwidth, degree of diagonal dominance, degree of sparsity, number of available processors, etc.) affect the solution strategy, the solver contains an adaptive layer to automatically select the optimum strategy at runtime. The solver is parallelized with MPI to take advantage of high-performance computing (HPC) clusters and other parallel architectures.
Solving banded linear systems plays a critical role in many applications. Banded systems frequently arise from reordering sparse matrices. In other instances, they are constructed as effective preconditioners to general sparse systems where they are solved via iterative methods. Existing parallel software using direct methods for banded matrices are mostly based on LU factorization. The Intel Adaptive Spike-Based Solver is based on a different decomposition method that balances communication overhead with arithmetic cost to achieve better scalability than other methods. The Intel Adaptive Spike-Based Solver offers HPC users a new and valuable tool for solving large banded systems.
The software can be downloaded from the product page:http://software.intel.com/en-us/articles/intel-adaptive-spike-based-solver.
- Superior performance over LU factorization
- Solution strategy adapts to matrix characteristics
- Parallel and scalable
- Fortran 90 and C compatibility
- Supports several compilers and MPI implementations
The Intel Adaptive Spike-Based Solver is a collaborative effort of Purdue University, the University of Massachusetts Amherst, and Intel Corporation.