ISC'12

June 17-21, 2012

Hamburg, Germany

Prof. Dr. Richard Murphy

Principal Member of Technical Staff, Sandia National Laboratories

Richard Murphy is a Principal Member of Technical Staff at Sandia National Laboratories and an affiliated faculty member at New Mexico State University. At Sandia, he is a leader in research and development of computer architecture, advanced memory systems, supercomputing systems for physics and data intensive problems. He helped to found a robust research program in memory systems at Sandia ranging from custom processors based on 3D Stacked DRAM to more commodity solutions enabling data movement in the memory system (scatter/gather, etc.). He helped to create and currently chairs the steering committee for the newly formed Graph 500 benchmark, which seeks better measurements for large-­‐scale data-­‐intensive applications. In 2007, he helped to start the DOE Institute for Advanced Architecture (IAA), and currently serves on the steering committee. He is also a member of the Alliance for Computing at Extreme Scale (ACES) design team, which was founded by Sandia and Los Alamos to deploy the Cielo system and NNSA’s next capability supercomputer. Murphy leads the X-­‐ caliber processor R&D project at Sandia, which is a DARPA UHPC funded project with key university and industry collaborators, is in the process of architecting an advanced core for large graph problems called ELVIS, and leads Sandia’s Exascale Grand Challenge (XGC). He has pursued several “first of a kind” projects in which Sandia’s fab is being used in the creation of advanced supercomputer prototypes. From 2000-­‐2002, Murphy was a project and technical lead at Sun Microsystems in the areas of Dynamic Reconfiguration and hardware resource management. While at Sun, he worked in all phases of product development, including the initial machine design process, first hardware delivery, product life-­‐cycle, and end-­‐of-­‐life. He additionally served as CTO of a startup company in the dot-­‐com era. Murphy was recognized HPCwire as a Person to Watch in 2011.

 
Events: Supercomputer Architectures for Data Intensive Computing
 
Contributions: New Developments for Ranking Supercomputers
Graph500 – Philosophy, Challenges & the State of the List
  
Large Memory Systems & Challenges
Large Memory Systems: An Applications View
  
BoFs 3 & 4
BoF 3: Fourth Graph500 List
  
Program may be subject to changes.