Asynchronous Iterative Algorithm for Computing Incomplete Factorizations on GPUs
نویسندگان
چکیده
This paper presents a GPU implementation of an asynchronous iterative algorithm for computing incomplete factorizations. Asynchronous algorithms, with their ability to tolerate memory latency, form an important class of algorithms for modern computer architectures. Our GPU implementation considers several non-traditional techniques that can be important for asynchronous algorithms to optimize convergence and data locality. These techniques include controlling the order in which variables are updated by controlling the order of execution of thread blocks, taking advantage of cache reuse between thread blocks, and managing the amount of parallelism to control the convergence of
منابع مشابه
Iterative Sparse Triangular Solves for Preconditioning
Sparse triangular solvers are typically parallelized using levelscheduling techniques, but parallel efficiency is poor on high-throughput architectures like GPUs. We propose using an iterative approach for solving sparse triangular systems when an approximation is suitable. This approach will not work for all problems, but can be successful for sparse triangular matrices arising from incomplete...
متن کاملSparse Approximate Inverse Preconditioners for Iterative Solvers on GPUs
For the solution of large systems of linear equations, iterative solvers with preconditioners are typically employed. However, the design of preconditioners for the black-box case, in which no additional information about the underlying problem is known, is very difficult. The most commonly employed method of incomplete LU factorizations is a serial algorithm and thus not well suited for the ma...
متن کاملDomain Overlap for Iterative Sparse Triangular Solves on GPUs
Iterative methods for solving sparse triangular systems are an attractive alternative to exact forward and backward substitution if an approximation of the solution is acceptable. On modern hardware, performance benefits are available as iterative methods allow for better parallelization. In this paper, we investigate how block-iterative triangular solves can benefit from using overlap. Because...
متن کاملDirect and Incomplete Cholesky Factorizations with Static Supernodes
Introduction Incomplete factorizations of sparse symmetric positive definite (SSPD) matrices have been used to generate preconditioners for various iterative solvers. These solvers generally use preconditioners derived from the matrix system, , in order to reduce the total number of iterations until convergence. In this report, we investigate the findings of ref. [1] on their method for computi...
متن کاملGPU-Accelerated Asynchronous Error Correction for Mixed Precision Iterative Refinement
In hardware-aware high performance computing, blockasynchronous iteration and mixed precision iterative refinement are two techniques that are applied to leverage the computing power of SIMD accelerators like GPUs. Although they use a very different approach for this purpose, they share the basic idea of compensating the convergence behaviour of an inferior numerical algorithm by a more efficie...
متن کامل