Student Contribution

SC Conference - Activity Details

Global Trees: A Framework for Linked Data Structures on Distributed Memory Parallel Systems

D. Brian Larkins  (Ohio State University)
James Dinan  (Ohio State University)
Sriram Krishnamoorthy  (Pacific Northwest National Laboratory)
Srinivasan Parthasarathy  (Ohio State University)
Atanas Rountev  (Ohio State University)
P. Sadayappan  (Ohio State University)
Papers Session
System Performance Optimization
Thursday,  04:00PM - 04:30PM
Room Ballroom F
This paper describes the Global Trees (GT) system which provides a multi-layered logical interface to a global view of distributed tree data structures, while providing scalable performance on distributed memory systems. The Global Trees system relies on coarse-grained data movement to increase locality and communication efficiency. We describe the design and implementation of GT, illustrate its use in the context of a well-known gravitational simulation application, and provide experimental results which validate this approach. The key benefits of using the this system include efficient shared-memory style programming of distributed trees, tree-specific optimizations for data access and computation, and the ability to customize many aspects of GT to provide highly tuned application performance.
The full paper can be found in the IEEE Xplore Digital Library and ACM Digital Library
   IEEE Computer Society  /  ACM     2 0   Y E A R S   -   U N L E A S H I N G   T H E   P O W E R   O F   H P C