Variable Precision Computing

Technical Report · 31 October 2019

DOI:https://doi.org/10.2172/1573151· OSTI ID:1573151

Hittinger, Jeffrey A. F. ^[1]; Lindstrom, P. G. ^[1]; Bhatia, H. ^[1]; Bremer, P. T. ^[1]; Copeland, D. M. ^[1]; Chand, K. K. ^[1]; Fox, A. L. ^[1]; Lloyd, G. S. ^[1]; Menon, H. ^[1]; Morrison, G. D. ^[1]; Osei-Kuffuor, D. ^[1]; Pinnow, N. T. ^[1]; Quinlan, D. J. ^[1]; Sanders, G. D. ^[1]; Schordan, M. ^[1]; Vanderbruggen, T. ^[1]; Hoang, D. ^[2]; Klacansky, P. ^[2]; Pascucci, V. ^[2]; Usher, W. ^[2] ; Lam, M. ^[3]; Moody, L. G. ^[3]; Diffenderfer, J. D. ^[4]; Metere, A. ^[5]; Yang, L. M. ^[6]

---

1. Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
2. Univ. of Utah, Salt Lake City, UT (United States)
3. James Madison Univ., Harrisonburg, VA (United States)
4. Univ. of Florida, Gainesville, FL (United States)
5. International Computer Science Institute (ICSI), Berkeley, CA (United States)
6. Univ. of Colorado, Boulder, CO (United States)

This report summarizes the activities and major accomplishments of the Variable Precision Computing Strategic Initiative project. The overarching goal of this project was to initiate and promote a new paradigm in High Performance Computing (HPC) that would fundamentally change how we represent and make use of representations of real numbers in finite precision and that would establish Lawrence Livermore National Laboratory (LLNL) as a leader in precision-related research for HPC. We pursued three integrated and concurrent research thrusts: new, more dynamic data representations to address data motion and capacity limitations; improved mixed precision algorithms to accelerate computation and to leverage new hardware; and new tools to help developers reason about precision and to automate code transformations. Within these thrust areas, we made significant advances in the use of compressed array data types in numerical calculations, in the theoretical justification, in improved performance, and in the prospects of hardware implementation; in the development of metrics to significantly reduce the storage needs of molecular dynamics simulations; in the development and analysis of efficient mixed-precision algorithms for time-dependent partial differential equations and for orthogonal factorization; and in the capability of tools that can analyze sensitivity of computed results to choices in precision and that can make automated code transformations based on such analysis. We established that there are definite opportunities to realize significant improvements in HPC above and beyond the 2x limits of traditional mixed precision using the technologies we have developed. The project produced numerous papers, presentations, and posters in important venues as well as open-source software contributions to the community, and many of the new capabilities have initiated or are being leveraged in ongoing projects.

• Research Organization: Lawrence Livermore National Laboratory (LLNL), Livermore, CA (United States)
• Sponsoring Organization: USDOE National Nuclear Security Administration (NNSA); USDOE Laboratory Directed Research and Development (LDRD) Program
• DOE Contract Number: AC52-07NA27344
• OSTI ID: 1573151
• Report Number(s): LLNL-TR-795750; 997173
• Country of Publication: United States
• Language: English

Related Subjects

97 MATHEMATICS AND COMPUTING