|PROPOSAL NUMBER:||06 S8.05-8900|
|SUBTOPIC TITLE:||Data Management and Visualization|
|PROPOSAL TITLE:||10-100 Gbps Offload NIC for WAN, NLR, Grid Computing|
SMALL BUSINESS CONCERN
(Firm Name, Mail Address, City/State/Zip, Phone)
SeaFire Micros, Inc.
39 Dodge St #319
Beverly, MA 01915-1705
PRINCIPAL INVESTIGATOR/PROJECT MANAGER
(Name, E-mail, Mail Address, City/State/Zip, Phone)
39 Dodge St #319
Beverly, MA 01915-1705
TECHNICAL ABSTRACT ( Limit 2000 characters, approximately 200 words)
This Phase 1 proposal addresses the 2006 NASA SBIR Research Topic S8.05 Science Data Management and Visualization (GSFC). The subtopic we address is Distributed Data Management and Access: Dynamically configurable high speed access to data distributed and shared over wide area high speed network environments. We propose a novel offload engine NIC, with network connections scaling at n x 10Gbps increments. The application is for present (10Gbps) and next generation NASA and commercial cluster and grid computing systems requiring ultra high-speed, low-cost offload engines with rates ranging from 10 through future rates of 100 Gbps full-duplex.
Successful completion of this Phase 1 project will prove feasibility of our innovation through architectural models, ultra high-speed logic simulations, and experimental results. Deliverables include system on chip core IP, initial firmware, and feasibility results showing UDP/IP from 10 - 100 Gbps full-duplex, and the maximum bandwidth for a variant of TCP/IP. We will also have the system design for our Phase 2 prototype system. Using a 10 Gbps NPU development system, we conduct a preliminary demonstration of offload engine performance.
POTENTIAL NASA COMMERCIAL APPLICATIONS ( Limit 1500 characters, approximately 150 words)
Our system would accommodate network bandwidth needs for host offload engines used within NASA Science Data Centers, Cluster, and Grid computing applications for present 10 Gbps full-duplex offload, and for future bandwidths scaled to 100 Gbps. Our novel offload engine NIC will be used for access to distributed and shared data applications including:
- National Lambda Rail (NLR) applications such as
a. Recent GSFC use of the NLR with 10 Gigabit Ethernet per wavelength.
b. Distributed services and data using Optiputer
c. Compatibility with GSFC networks and Science Data Centers that
ultimately tie into NLR.
- Enabling NASA Applications across Heterogeneous High Performance Networks, where multiple 10Gbps connections to DRAGON are used.
- NCCS and other GSFC Science Data facilities with CPU's linked to systems including multiple 10-GE ports, and to networks aggregated in multiples of 10GigE
- Supporting new protocol variations of UDP and TCP
POTENTIAL NON-NASA COMMERCIAL APPLICATIONS ( Limit 1500 characters, approximately 150 words)
Successful completion of this project, through Phase 3, would allow us to provide an offload engine NIC product suitable for the commercial sector Grid computing market. The resulting system-on-chip intellectual property (core IP) could be marketed to offload engine manufacturers. The IP would save existing offload engine firms design time, design cost, and provide excellent time-to-market. Additional customers would include motherboard and systems manufacturers that require offload engine NIC's or system-on-chips for their motherboards, such as Hewlett-Packard, IBM, and Dell. IP distribution channels and sales could also be through system-on-chip providers. National laboratories requiring ultra high-speed cluster and grid computing offload engines could also benefit from NASA's investment in our innovation. Raytheon has provided a letter of endorsement, for use in their future Grid and radar systems applications.
|NASA's technology taxonomy has been developed by the SBIR-STTR program to disseminate awareness of proposed and awarded R/R&D in the agency. It is a listing of over 100 technologies, sorted into broad categories, of interest to NASA.|
TECHNOLOGY TAXONOMY MAPPING
Architectures and Networks
Computer System Architectures
Software Tools for Distributed Analysis and Simulation