Solving sparse linear systems with approximate inverse preconditioners on analog devices
Vasileios Kalantzis, Anshul Gupta, Lior Horesh, Tomasz Nowicki, Mark S. Squillante, Chai Wah Wu, Tayfun Gokmen, Haim Avron
Abstract
Sparse linear system solvers are computationally expensive kernels that lie at the heart of numerous applications. This paper proposes a preconditioning framework that combines approximate inverses with stationary iterations to substantially reduce the time and energy requirements of this task by utilizing a hybrid architecture that combines conventional digital microprocessors with analog crossbar array accelerators. Our analysis and experiments with a simulator for analog hardware show that an order of magnitude speedup is readily attainable despite the noise in analog computations.
Topics & Concepts
SpeedupComputer scienceComputationAnalog computerCrossbar switchLinear systemSparse matrixNoise (video)InverseComputer engineeringComputational scienceElectronic engineeringParallel computingAlgorithmArtificial intelligenceMathematicsElectrical engineeringEngineeringImage (mathematics)PhysicsGaussianMathematical analysisTelecommunicationsGeometryQuantum mechanicsMatrix Theory and AlgorithmsElectromagnetic Scattering and AnalysisLow-power high-performance VLSI design