This shows you the differences between two versions of the page.

Link to this comparison view

seminars:seminar_10_18_03 [2017/09/20 22:02] (current)
Line 1: Line 1:
 +====== Increasing Efficiency and Scalability with Tunable Error Protection and Memory Access Granularity ======
 +Thursday March 18, 2010\\
 +12:00 noon\\
 +**[[http://​users.ece.utexas.edu/​~merez/​|Mattan Erez]]**\\
 +UT Austin ECE\\
 +I will present my group'​s vision and initial work towards enabling more
 +efficient and scalable systems through dynamic and flexible cooperative
 +error protection and variable granularity memory systems. Flexible
 +protection recognizes that different computations and data require different
 +degrees of protection, and hence different amounts of resources for
 +acceptable execution outcomes. ​ Variable granularity memory systems relax
 +the trend of ever-increasing access granularities and improve utilization of
 +scarce memory bandwidth and power resources. While the two topics may seem
 +unrelated, they are in fact closely tied to one another and share many of
 +the same enabling mechanisms.
 +I will present motivating examples for when to apply cooperative protection
 +and variable granularity,​ as well as their overall potential given current
 +system design trends. I will then discuss our initial set of hardware
 +mechanisms, and explain why dynamic protection and access granularity are
 +intimately related. These mechanisms involve a range of tradeoffs with
 +respect to redundancy, storage capacity, error-tolerance complexity, and
 +expected performance. I will conclude by discussing how to best expose the
 +hardware flexibility to the programmer and software and a framework that
 +will enable exploring the tradeoff space.
 +Mattan Erez is an Assistant Professor in the Department of Electrical and
 +Computer Engineering at the University of Texas at Austin. Mattan received a
 +B.Sc. in Electrical Engineering and a B.A. in Physics from the Technion,
 +Israel Institute of Technology in 1999. He subsequently received his M.S and
 +Ph.D. in Electrical Engineering from Stanford University in 2002 and 2007
 +respectively. His experience includes working as a computer architect in the
 +Israeli Processor Architecture Research team, Intel Corporation. As a Ph.D.
 +candidate at Stanford University he was the student leader of the Merrimac
 +Stanford Streaming Supercomputer project, where his work spanned the entire
 +system from microarchitecture to the Brook and Sequoia programming models.
 +Mattan is continuing this whole-system approach at UT Austin, where his
 +research focuses on improving performance and efficiency through advances in
 +processor architecture and programming models and techniques.
 +**[[seminars| Back to the seminar page]]**