An Analog Preconditioner for Solving Linear Systems [Slides]

Conference · 03 March 2021

DOI:https://doi.org/10.2172/1848047· OSTI ID:1848047

Feinberg, Benjamin ^[1]; Wong, Ryan Ming ^[2]; Xiao, Tianyao Patrick ^[1]; Bennett, Christopher H. ^[1]; Rohan, Jacob Nathanael ^[3]; Boman, Erik Gunnar ^[1]; Marinella, Matthew J. ^[1]; Agarwal, Sapan ^[1]; Ipek, Engin ^[4]

---

1. Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
2. Univ. of Rochester, NY (United States)
3. Univ. of Texas, Austin, TX (United States)
4. Qualcomm Inc., San Diego, CA (United States)

This presentation concludes in situ computation enables new approaches to linear algebra problems which can be both more effective and more efficient as compared to conventional digital systems. Preconditioning is well-suited to analog computation due to the tolerance for approximate solutions. When combined with prior work on in situ MVM for scientific computing, analog preconditioning can enable significant speedups for important linear algebra applications.

• Research Organization: Sandia National Lab. (SNL-NM), Albuquerque, NM (United States)
• Sponsoring Organization: USDOE National Nuclear Security Administration (NNSA)
• DOE Contract Number: NA0003525
• OSTI ID: 1848047
• Report Number(s): SAND2021-1603C; 694311
• Resource Relation: Conference: 27.IEEE International Symposium on High-Performance Computer Architecture (HPCA-27), Held Virtually, Seoul (South Korea), 27 Feb - 3 Mar 2021; Related Information: https://hpca-conf.org/2021/main-program/
• Country of Publication: United States
• Language: English

Related Subjects

97 MATHEMATICS AND COMPUTING