Redefining Analytics for Small High-Performance Computing Clusters
Abstract
The problem is that existing parallel analytical systems, such as Shark and Hadoop, are built such that they do not effectively leverage the benefits of current hardware trends (high performance RDMA capable networks and high-end many-core machines with large amounts of main memory). Existing parallel analytical systems appear to target huge cloud deployments with cheap but low-end machines connected via high-latency low-bandwidth networks. Since most infrastructures do not operate with this archaic hardware, the objective of this proposal is to utilize Small High-Performance Computing (SHPC) clusters to advance statistical machine learning techniques and agile analytics.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Mar 23, 2016
- Source ID
- FA95501510144
Entities
People
- Tim Kraska
Organizations
- Air Force Office of Scientific Research
- Brown University
- United States Air Force