Skip Navigation

Colloquium Details

Remote Memory Access Programming: Faster Parallel Computing Without Messages

Author:Torsten Hoefler, Assistant Professor ETH Zurich, Switzerland
Date:June 15, 2015
Time:10:00
Location:220 Deschutes

Abstract

Remote memory access (RMA) or partitioned global address space programming offers abstractions to coordinate directly accessible distributed memory domains. The presentation will start with an analysis of established technologies (cache-coherence), continue to describe RMA programming abstractions and example codes, and conclude with proposals for the future of RMA. In the first part, we will use an intuitive performance model to briefly demonstrate the complexities of cache-coherent systems with regards to performance tuning. We will exemplify these findings with results on Intel's Xeon Phi and Sandy Bridge CPUs where we speed up communication algorithms by up to 4.3x. We continue by showing how RMA programming simplifies the design and tuning and introduce MPI-3's RMA semantics as a particuler example.

We discuss our reference implementation for Cray machines foMPI and demonstrate results with up to half a million processes. We conclude the talk by addressing producer-consumer synchronizations in task-based runtime environments and the new proposal of notified access. Overall, we advocate RMA as a potential programming model for scalable systems ranging from single-die multicores to large-scale supercomputers.

Biography

Torsten is an Assistant Professor of Computer Science at ETH Zürich, Switzerland. Before joining ETH, he led the performance modeling and simulation efforts of parallel petascale applications for the NSF-funded Blue Waters project at NCSA/UIUC. He is also a key member of the Message Passing Interface (MPI) Forum where he chairs the “Collective Operations and Topologies" working group.

Torsten won best paper awards at the ACM/IEEE Supercomputing Conference SC10, SC13, SC14, EuroMPI 2013, IPDPS 2015, and other conferences. He published numerous peer-reviewed scientific conference and journal articles and authored chapters of the MPI-2.2 and MPI-3.0 standards. His research interests revolve around the central topic of "Performance-centric Software Development" and include scalable networks, parallel programming techniques, and performance modeling. Additional information about Torsten can be found on his homepage at: htor.inf.ethz.ch.