Efficiency of Reproducible Level 1 BLAS

Chemseddine Chohra, Philippe Langlois, David Parello · Lecture notes in computer science · 2016

Numerical reproducibility failures appear in massively parallel floating-point computations. One way to guarantee this reproducibility is to extend the IEEE-754 correct rounding to larger computing sequences, e.g. to the BLAS. Is the extra cost for numerical reproducibility acceptable in practice? We present solutions and experiments for the level 1 BLAS and we conclude about their efficiency. These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Read the paper · More papers on PaperTik