NAG Library for SMP & multicore: Performance Examples
The routines in the NAG Library that benefit from SMP parallelism include problem solvers in the areas of: Dense and Sparse Linear Algebra; Fast Fourier Transforms; Random Number Generators; Quadrature; Partial Differential Equations; Interpolation; Curve and Surface Fitting; Correlation and Regression Analysis; Multivariate Methods; Time Series Analysis; Financial Option Pricing; Global Optimization and Wavelet Transforms.
The examples below, illustrate how the performance scales on multiple cores. They are drawn from Library Chapters that cover Correlation and Regression Analysis, Wavelet Transforms and Large Scale Linear Systems.
The sparse matrix examples are provided to illustrate the use of the same routine for solving different types and size of problem. For the particular problems chosen the performance levels out above 16 cores.
Hardware platform for these results:
24 Core Machine comprising: Two AMD Opteron 6174 processors. Each processor has 12 cores running at 2.2 GHz.
Note: Run times level out beyond 16 for the following problem type.
Note: Whilst a significant number of routines in the NAG Library for SMP and multicore exhibit improved runtime performance, compared to the equivalent routine in the NAG Fortran Library, this is not always the case. To determine the speedup you should consult the document Tuned and Enhanced Routines in the NAG Library for SMP & Multicore and read whether the routine you intend to use benefits, either directly or indirectly, from the use of multiple processors. .