Speaker: Madan Musuvathi
Affiliation: Microsoft Research
Parallelizing Seemingly-Sequential Computations
ABSTRACT: Important applications, such as machine learning and log processing, involve iterating over a large data set with loop-carried dependences across iterations. As such, these computations are not embarrassingly parallel. In this talk, I will describe symbolic parallelism, a general methodology for parallelizing dependent computations. The basic idea is to break dependences using efficient symbolic reasoning.
This work is primarily motivated by the need to extract parallelism from data processing queries that combine standard relational operations (such as map, reduce, filter) with non-relational application-specific code. By applying symbolic parallelism to such code and exposing the resulting parallelism to the underlying query optimizer, we obtain 2-4 orders of magnitude performance improvements over Hadoop and SQL Server.
BIO: Madan Musuvathi is a Principal Researcher at Microsoft Research working in the intersection of programming languages and systems, with specific focus on concurrency and parallelism. His interests span program analysis, systems, model checking, verification, and theorem proving. His research has led to several tools that improve the lives of software developers both at Microsoft and at other companies. He received his Ph.D. from Stanford University in 2004.
Hosted by Todd Millstein
REFRESHMENTS at 3:45 pm, SPEAKER at 4:15 pm
Date(s) - Dec 01, 2015
4:15 pm - 5:45 pm