The Utpal Banerjee Distinguished Lecture in High Performance Computing: Charles Leiserson, "Setting a Course for Post-Moore Software Performance"

Abstract: 
Software performance engineering (SPE) is the science and art of making code run fast or otherwise limiting its consumption of resources, such as power, memory footprint, network utilization, file IO’s, response time, etc. SPE encompasses parallel computing, but it also includes other techniques, such as caching, vectorization, algorithms, bit tricks, loop unrolling, compiler-switch selection, tailoring code to the architecture, exploiting sparsity, changing data representation, metaprogramming, etc.  Historically, gains in performance from miniaturization, codified in Moore's Law, relieved programmers from the burden of making software run fast and learning SPE techniques. I will explain why the end of Moore’s Law has made SPE a critical technical skill. Since SPE is neither extensively researched nor widely taught in the universities, however, it risks devolving into an unstructured collection of ad hoc tricks.  Now is the time to establish SPE as a science-based discipline, alongside traditional areas of computer science.

Bio:
Charles E. Leiserson is Edwin Sibley Webster Professor of Computer Science and Engineering in MIT’s Department of Electrical Engineering and Computer Science (EECS) and a member and former Associate Director of MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL).  He received a B.S. in computer science and mathematics from Yale University in 1975 and a Ph.D. in computer science from Carnegie Mellon University in 1981. He currently serves as the MIT Faculty Director of the Department of the Air Force-MIT AI Accelerator and leads its Fast AI project. His award-winning research on algorithms, parallel computing, and software performance engineering has been widely deployed in industry.  He held the position of Director of System Architecture for the MIT-spinoff Akamai Technologies, and he founded Cilk Arts, Inc., a multicore-software start-up acquired by Intel. He was the network architect for the Connection Machine CM-5, the world’s most powerful computer in 1993. He coauthored the influential textbook Introduction to Algorithms, which has sold over one million copies. Leiserson is a Fellow of four professional societies—ACM, AAAS, SIAM, and IEEE—and he is a member of the National Academy of Engineering.