Improving the Performance of a Verified Linear System Solver Using Optimized Libraries and Parallel Computation