EECS Seminar: How Much Time, Energy and Power Does an Algorithm Need?
Abstract: Given an algorithm, can we estimate or bound the amount of physical energy (joules) or power (watts) it might require, in the same way that we do for time and storage? Models of these physical measures of performance are of increasing importance in nearly every class of computing device, from embedded mobile systems to power-constrained supercomputers. Armed with such models, we can try to answer many interesting questions. For instance, can algorithmic knobs be used to control energy or power as the algorithm runs? How might systems be better balanced in energy or power for certain classes of algorithms? This talk covers general ideas of what such analyses and models might look like, giving both theoretical predictions and early empirical validation of our algorithmic energy and power models on real software and systems.
Bio: Rich Vuduc is an associate professor in the School of Computational Science and Engineering at the Georgia Institute of Technology (Georgia Tech). His research lab, the HPC Garage (@hpcgarage), is interested in performance analysis and performance engineering. He has received a DARPA Computer Science Study Group grant, an NSF CAREER award, a collaborative Gordon Bell Prize in 2010, Lockheed Martin’s Award for Excellence in Teaching (2013, and Best Paper Awards at the SIAM Conference on Data Mining (SDM, 2012) and the IEEE Parallel and Distributed Processing Symposium (IPDPS, 2015), among others. He also serves as his department’s associate chair and director of its graduate programs. Vuduc was elected vice president of the SIAM Activity Group on Supercomputing (2016-17); and serves as associate editor of both the International Journal of High-Performance Computing Applications (IJHPCA) and IEEE Transactions on Parallel and Distributed Systems (TPDS). He received his Ph.D. in computer science from the University of California, Berkeley, and was a postdoctoral scholar at the Lawrence Livermore National Laboratory.