Hybrid Photonic Iterative Solver for Large-Scale Systems of Equations

Technology #11650

Key Points

  • A hybrid analog and digital computational system for solving large-scale systems of equations and optimization problems
  • Dynamic fixed-point format overcomes the error stagnation caused by the limited precision of fixed-point computing
  • Compatible with the latest hardware advancement in high-performance computing

Abstract

The University of Central Florida invention relates to a method of solving a linear equation system for use in matrix inversion, image reconstruction, high–dimensional inverse problems, numerical eigen problems, and ab initio ray-tracing engines.