[Colloquium] Thomas Sterling Talk, 1/19/23

[Sandy Quarles]

Department of Computer Science Seminar

Thomas Sterling 

Professor of Intelligent Systems Engineering School of Informatics, Computing, and Engineering

Indiana University 

Thursday, January 19th
1:00pm - 2:00pm
Crerar Library 298

Active Memory Architecture: memory-centric non von Neumann Dynamic Graph Processing

Abstract:  A revolution in scalable computer system design is emerging in the domains of Machine Learning, Data Analytics, and Graph Processing. Leaders in the commercial field (e.g., Google) and energetic startups (e.g., Graphcore) with combined assets in the billions of dollars are pushing the boundaries of these perceived market opportunities through R&D in computer architecture, API models, and software. The new IARPA AGILE research program has been created to advance the state-of-the-art beyond these near-term commercial techniques with the University of Chicago and Indiana University among the recently selected performers. The 3-year AGILE program, recently undertaken, seeks major innovations in architecture, runtime software, and parallel algorithms to achieve one to two orders of magnitude performance gains within this decade for capabilities in graph-data and machine intelligence. The experimental Active Memory Architecture is an innovative memory centric non von Neumann system concept devised to circumvent legacy bottlenecks of traditional von Neumann-based cores and develop smart memories to accelerate primitive operations central to graph execution including irregular time-varying data structures. This seminar will describe work-in-progress for the innovative Active Memory Architecture (AMA) to meet these challenging goals for Machine Intelligence in its diversity of forms. AMA is conceived to deliver dramatic capability gains in dynamic graph computation and is combined with the proposed system-wide Pilot-CRIS scalable structure for full-graph capacity handling. Qualitative justification will be discussed to motivate adoption of the AMA conceptual strategy. Questions and comments will be welcome from the audience throughout the technical presentation.
Bio:Thomas Sterling is a Full Professor of Intelligent Systems Engineering at Indiana University (IU) serving as Director of the AI Computing Systems Laboratory at IU’s Luddy School of Informatics, Computing, and Engineering. Since receiving his Ph.D from MIT as a Hertz Fellow in 1984, Dr. Sterling has engaged in applied research in parallel computing system structures, semantics, and operation in industry, government labs, and academia. Professional affiliations have included Harris Corp., IDA Supercomputing Research Center, NASA (GSFC, JPL), Un. of Maryland, Caltech, and LSU. Dr. Sterling is best known as the "father of Beowulf" for his pioneering research in commodity/Linux cluster computing for which he shared the Gordon Bell Prize in 1997. His current research is associated with innovative extreme scale computing through memory-centric non von Neumann architecture concepts to accelerate dynamic graph processing for AI including ML. In 2018, he co-founded the new tech company, Simultac LLC, and serves as its President and Chief Scientist. Dr. Sterling was the recipient of the 2013 Vanguard Award and is a Fellow of the AAAS. He has been selected this year to be inducted in the Space Technologies Hall of Fame. He is the co-author of seven books and holds six patents. Most recently, he co-authored the introductory textbook, “High Performance Computing”, published by Morgan-Kaufmann in 2018 which is going into 2nd edition. 





Host:  Andrew A. Chien




-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20230109/d4ec8f09/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: PastedGraphic-1.tiff
Type: image/tiff
Size: 156420 bytes
Desc: not available
URL: <http://mailman.cs.uchicago.edu/pipermail/colloquium/attachments/20230109/d4ec8f09/attachment-0001.tiff>