The documents distributed by this server have been provided by the contributing authors as a means to ensure timely dissemination of scholarly and technical work on a noncommercial basis. Copyright and all rights therein are maintained by the authors or by other copyright holders, notwithstanding that they have offered their works here electronically. It is understood that all persons copying this information will adhere to the terms and constraints invoked by each author's copyright. These works may not be reposted without the explicit permission of the copyright holder.
Publications of SPCL
|Performance-oriented Parallel Programming Integrating Hardware, Middleware, and Applications|
(Presentation - presented in Savannah, GA, USA, Feb. 2012)
SIAM SIAG/SC Junior Scientist Award Lecture
AbstractParallel programming is hard, optimizing parallel programming is even harder, and writing optimal parallel programs is nearly impossible. In this talk, we discuss how optimizing communication in parallel programs routinely requires to deal with low-level system details. We show portable abstractions that enable transparent optimizations but require advanced techniques in the lower layers. We conclude that scaling to larger machines demands an holistic approach to integrate hardware, middleware, and application software to develop performance-portable parallel programs.