You are here

Operating System Mechanisms for Many-Core Systems

Award Information
Agency: Department of Defense
Branch: Air Force
Contract: FA8750-13-C-0167
Agency Tracking Number: O2-1370
Amount: $749,992.00
Phase: Phase II
Program: STTR
Solicitation Topic Code: OSD11-T04
Solicitation Number: 2011.B
Timeline
Solicitation Year: 2011
Award Year: 2013
Award Start Date (Proposal Award Date): 2013-05-30
Award End Date (Contract End Date): 2015-05-29
Small Business Information
1050 W NASA Blvd Suite 155
Melbourne, FL -
United States
DUNS: 038379579
HUBZone Owned: No
Woman Owned: No
Socially and Economically Disadvantaged: No
Principal Investigator
 Bruce McQueary
 Vice President
 (321) 591-7371
 bmcqueary@securboration.com
Business Contact
 Lynn Lehman
Title: CEO
Phone: (919) 244-3946
Email: contracts@securboration.com
Research Institution
 North Carolina State University
 John Chaffee
 
2701 Sullivan Drive Suite 240 (Campus Box 7514)
Raleigh, NC 27695-7514
United States

 (919) 515-2444
 Nonprofit College or University
Abstract

In the Phase I portion of this STTR, Securboration and renowned multicore expert Dr. Frank Mueller from North Carolina State University designed, developed, and benchmarked the proof-of-concept Pico-kernel Adaptive and Scalable Operating-system (PICASO) for many-core architectures. The Securboration Team took a scientific, experimentation-based approach to identifying and resolving shortcomings with existing Operating Systems (OS) design when applied to many (or multi)-core architectures. Based on these findings, the Team developed an innovative solution that uses sets of coordinated, interoperating micro-kernels and pico-kernels that communicate via distributed message passing. Each kernel resides on an individual core, and in a hierarchical fashion the micro-kernels'manage'adjacent sets of pico-kernels. The Team ran PICASO through the standardized NASA Advanced Supercomputing Parallel Benchmark (NPB) suite and compared the results with other multicore OS implementations based on shared memory and message passing. In comparison, PICASO showed significant performance increases in all metrics with minimal overhead as the number of cores scaled from two to the maximum available (sixty-four). PICASO's ability to maximize multicore performance has significant commercial potential for companies such as Facebook and Google who are seeking to optimize their performance cost.

* Information listed above is at the time of submission. *

US Flag An Official Website of the United States Government