|
Laboratory Overview
Founded in 1993 by Dr. Alan D. George, the
High-performance Computing and Simulation (HCS) Research Laboratory
focuses on research in advanced computer architectures, networks, systems,
services, and applications for critical applications in reconfigurable,
parallel, distributed, and fault-tolerant computing. The lab is
headquartered in the ECE Department
at the University of Florida
(HCS-Gainesville), with a partner lab site at the FAMU-FSU College of Engineering
(HCS-Tallahassee). Both sites house key facilities linked by Florida Lambda Rail. The lab has been
cited by the NSA as a Research Center of Excellence in
High-Performance Computing and Networking.
Activities focus on core areas in high-performance computer
engineering whose contributions are critical for scalable,
high-performance, dependable, and secure communications and
computations far into the new century. Researchers address key issues
that span the entire spectrum, from low-level hardware to
grand-challenge applications, in a manner that emphasizes both
theoretical and applied research to bring to fruition new concepts,
models, techniques, and tools.
NSF Center for High-performance Reconfigurable Computing
After more than two years in development, the NSF Center for
High-Performance Reconfigurable Computing (CHREC, pronounced "shreck") is now
operational. This new national research center and consortium was
selected and awarded in September 2006 by the I/UCRC program at the National Science Foundation and began
research with a kickoff workshop in December 2006 then operational
status in January 2007. See www.chrec.org for more information.
Research Projects Research in the lab is supported by a strong
cadre of federal sponsors including DOD, NASA, NSF, and NIH, and
leading industry sponsors including Honeywell Space and Cray.
Activities in the lab are undertaken by several research
groups in areas including advanced space computing, reconfigurable
computing, parallel performance optimization with UPC, high-performance
computing applications, high-speed networks, and mission assurance for
HPC.Current projects involving lab members include the following:
- Dependable Multiprocessor: toward the first supercomputer in space (in partnership with Honeywell, funded by NASA)
- PPW: Performance analysis and realization for UPC and SHMEM (funded by DOD)
- RapidIO networks, systems, and applications for high-performance embedded computing in space (funded by Honeywell)
- Dolphin: WDM-based local-area networks for advanced avionics systems (funded by NAVAIR)
- Mission assurance strategies for resilient high-performance computing (funded by DOD)
- Energy efficiency for edge nodes on the Internet (in partnership with USF, funded by NSF)
- UltraLight: an ultrascale information system for data-intensive research (in partnership with Caltech, funded by NSF)
- iVDGL: the international virtual data grid laboratory (funded by NSF)
- Reconfigurable supercomputing: applications and services on the Cray XD1 (supported by Cray)
Some of many prior projects of note include:
- Hardware-reconfigurable, network-attached architecture for mission-critical systems (funded by DOD)
- High-performance algorithms and applications for computational biomechanics and optimization (funded by NIH)
- High-performance algorithms, architectures, and services for advanced sonar signal processing (funded by ONR)
- GEMS: highly scalable and robust monitoring service based on gossiping (funded by Sandia)
- ISE: integrated simulation environment for rapid virtual prototyping of parallel systems and applications (funded by DOD)
- LION: library for integrated optical networking (funded by Rockwell Collins and ML Design Technologies)
- QoS in Gigabit Ethernet for advanced avionics networks (funded by Rockwell Collins)
FacilitiesHoused in over 3000 square feet of laboratory space,
facilities in the HCS Lab are among the finest in the nation for
research in high-performance computer engineering. A computational grid is featured and consists of
12 Linux clusters with some 500 processors, the newest being four
Opteron and Xeon clusters. A broad range of high-speed networking testbeds are outfitted in these
clusters, including the latest Quadrics, DDR and SDR InfiniBand, 10
Gigabit Ethernet, and Gigabit Ethernet technologies. Some of the
servers are outfitted with powerful resources
for FPGA-based reconfigurable computing in the form of PCI and PCI-X
cards. Of special note is our new RC cluster known as Delta, our XD1 system on loan from Cray, and our new XD1000 Opteron/FPGA
multiprocessor.
Computer Engineering The focus of the HCS Lab is in the field of
computer engineering (in particular, high-performance computer
engineering). With the advent of digital computing and communications
systems during the past half-century, and their explosive growth into
prominence in every facet of our lives during the past several decades,
computer engineering has come to the forefront as one of the most
important areas of engineering research for current and future societies.
From its theoretical and historical roots in electrical engineering,
computer engineering has grown to encompass a wide variety of research
studies in both computer engineering infrastructure and applications,
spanning concepts to implementations, in terms of hardware and software.
Today, virtually every field of human endeavor, from aeronautical
engineering to medicine, has become increasingly dependent upon research
and technology advances in computer engineering.
|
 |
![[the lab]](/img/b_lab.gif)
![[projects]](/img/b_projects.gif)
![[people]](/img/b_people.gif)
