# Final report for “Extreme-scale Algorithms and Solver Resilience”

## Abstract

This is a joint project with principal investigators at Oak Ridge National Laboratory, Sandia National Laboratories, the University of California at Berkeley, and the University of Tennessee. Our part of the project involves developing performance models for highly scalable algorithms and the development of latency tolerant iterative methods. During this project, we extended our performance models for the Multigrid method for solving large systems of linear equations and conducted experiments with highly scalable variants of conjugate gradient methods that avoid blocking synchronization. In addition, we worked with the other members of the project on alternative techniques for resilience and reproducibility. We also presented an alternative approach for reproducible dot-products in parallel computations that performs almost as well as the conventional approach by separating the order of computation from the details of the decomposition of vectors across the processes.

- Authors:

- Univ. of Illinois, Urbana-Champaign, IL (United States)

- Publication Date:

- Research Org.:
- Univ. of Illinois, Urbana-Champaign, IL (United States)

- Sponsoring Org.:
- USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR) (SC-21)

- OSTI Identifier:
- 1411789

- Report Number(s):
- Final Report: DOE-UIUC-10049

- DOE Contract Number:
- SC0010049

- Resource Type:
- Technical Report

- Country of Publication:
- United States

- Language:
- English

- Subject:
- 97 MATHEMATICS AND COMPUTING; Multigrid; Conjugate Gradient; reproducibility; fault tolerance

### Citation Formats

```
Gropp, William Douglas.
```*Final report for “Extreme-scale Algorithms and Solver Resilience”*. United States: N. p., 2017.
Web. doi:10.2172/1411789.

```
Gropp, William Douglas.
```*Final report for “Extreme-scale Algorithms and Solver Resilience”*. United States. doi:10.2172/1411789.

```
Gropp, William Douglas. Fri .
"Final report for “Extreme-scale Algorithms and Solver Resilience”". United States.
doi:10.2172/1411789. https://www.osti.gov/servlets/purl/1411789.
```

```
@article{osti_1411789,
```

title = {Final report for “Extreme-scale Algorithms and Solver Resilience”},

author = {Gropp, William Douglas},

abstractNote = {This is a joint project with principal investigators at Oak Ridge National Laboratory, Sandia National Laboratories, the University of California at Berkeley, and the University of Tennessee. Our part of the project involves developing performance models for highly scalable algorithms and the development of latency tolerant iterative methods. During this project, we extended our performance models for the Multigrid method for solving large systems of linear equations and conducted experiments with highly scalable variants of conjugate gradient methods that avoid blocking synchronization. In addition, we worked with the other members of the project on alternative techniques for resilience and reproducibility. We also presented an alternative approach for reproducible dot-products in parallel computations that performs almost as well as the conventional approach by separating the order of computation from the details of the decomposition of vectors across the processes.},

doi = {10.2172/1411789},

journal = {},

number = ,

volume = ,

place = {United States},

year = {Fri Jun 30 00:00:00 EDT 2017},

month = {Fri Jun 30 00:00:00 EDT 2017}

}