Litcius/Paper detail

Eliminating Iterations of Iterative Methods

Tao Song, Xiaoming Chen, Yinhe Han

202132 citationsDOIOpen Access PDF

Abstract

The sparse linear solver is an important component in lots of scientific computing applications. For large-scale sparse linear systems, general-purpose processors such as CPUs and GPUs are facing challenges of high time complexity and massive data movements between processors and main memories. This work utilizes the ability of in-situ analog computing of RRAMs and builds an RRAMbased accelerator for iterative linear solvers.We first propose a basic principle of mapping iterative solvers onto RRAM-based crossbar arrays. The proposed principle eliminates not only the iterations but also the convergence condition. Based on the principle, we propose a scalable architecture that can solve large-scale sparse matrices in O(1) time complexity. Compared with a massively parallel iterative solver on GPU, our accelerator shows 100× higher performance and 1000× energy reduction. If the solution obtained by our accelerator is used as the seed for a further refinement on GPU, about 35% of the solving time and energy consumption can be saved compared with a pure GPU solving process.

Topics & Concepts

Computer scienceMassively parallelParallel computingSolverComputational scienceScalabilityIterative methodCUDASparse matrixConvergence (economics)Linear systemAlgorithmMathematicsMathematical analysisQuantum mechanicsProgramming languageGaussianEconomicsEconomic growthDatabasePhysicsAdvanced Memory and Neural ComputingFerroelectric and Negative Capacitance DevicesGraphene research and applications
Eliminating Iterations of Iterative Methods | Litcius