SC19 Proceedings

The International Conference for High Performance Computing, Networking, Storage, and Analysis

Poster 122: Project 38: Accelerating Architecture Innovation into Fieldable Extreme-Scale Systems (A Cross-Agency Effort)

Authors: John Shalf (Lawrence Berkeley National Laboratory), Dilip Vasudevan (Lawrence Berkeley National Laboratory), David Donofrio (Lawrence Berkeley National Laboratory), Anastasia Butko (Lawrence Berkeley National Laboratory), Andrew Chien (University of Chicago), Yuanwei Fang (University of Chicago), Arjun Rawal (University of Chicago), Chen Zou (University of Chicago), Raymond Bair (Argonne National Laboratory), Kristopher Keipert (Argonne National Laboratory), Arun Rodriguez (Sandia National Laboratories), Maya Gokhale (Lawrence Livermore National Laboratory), Scott Lloyd (Lawrence Livermore National Laboratory), Xiaochen Guo (Lehigh University), Yuan Zeng (Lehigh University)

Abstract: Accelerating technology disruptions and architectural change create growing opportunities and urgency to reduce the latency in for new architectural innovations to be deployed in extreme scale systems. We are exploring new architectural features that improve memory system performance including word-wise scratchpad memory, a flexible Recode engine, hardware message queues, and the data rearrangement engine (DRE). Performance results are promising yielding as much as 20x benefit. Project 38 is a cross-agency effort undertaken by the US Department of Energy (DOE) and Department of Defense (DoD).

Best Poster Finalist (BP): no

Poster: PDF
Poster summary: PDF

Back to Poster Archive Listing