A5052975541- Parallel Computing and Optimization Techniques
- Distributed and Parallel Computing Systems
- Advanced Data Storage Technologies
- Interconnection Networks and Systems
- Embedded Systems Design Techniques
- Cloud Computing and Resource Management
- Distributed systems and fault tolerance
- Scientific Computing and Data Management
- Software Testing and Debugging Techniques
- Software System Performance and Reliability
- Simulation Techniques and Applications
- Advanced Database Systems and Queries
- Formal Methods in Verification
- Multimedia Communication and Technology
- Energy Efficient Wireless Sensor Networks
- Cellular Automata and Applications
- Matrix Theory and Algorithms
- Scheduling and Optimization Algorithms
- Software Engineering Research
- Radiation Detection and Scintillator Technologies
- Mobile Agent-Based Network Management
- Software-Defined Networks and 5G
- VLSI and FPGA Design Techniques
- Neural Networks and Applications
- Advanced Malware Detection Techniques
Oak Ridge National Laboratory
2010-2022
Institute for the Future
2022
University of Arizona
2022
University of Tennessee at Knoxville
2011
University of New Mexico
1999-2009
United States Department of Commerce
1999
Sandia National Laboratories
1993-1999
Sandia National Laboratories California
1997
Los Alamos National Laboratory
1990-1991
Over the last 20 years, open-source community has provided more and software on which world’s high-performance computing systems depend for performance productivity. The invested millions of dollars years effort to build key components. However, although investments in these separate elements have been tremendously valuable, a great deal productivity also lost because lack planning, coordination, integration technologies necessary make them work together smoothly efficiently, both within...
The Program Dependence Web (PDW) is a program representation that can be directly interpreted using control-, data-, or demand-driven models of execution. A PDW combines single-assignment version the with explicit operators manage flow data values. viewed as an augmented Graph. Translation to provides basis for projects compile Fortran onto dynamic dataflow architectures and simulators. second application construction various compositional semantics dependence graphs.
In any assessment of potential terrorist attacks, the nuclear threat takes center stage. Although weapons-grade materials arc heavily guarded, a plausible scenario involves terrorists detonating simple radiological dispersion device (ROD) capable broadcasting nonfissile but highly radioactive particles over densely populated area. most cases, motor vehicle has to transport and its payload commonly known as "dirty bomb" - target destination. As final defense against such weapon, select...
This paper describes the evolution of Portals message passing architecture and programming interface from its initial development on tightly-coupled massively parallel platforms to current implementation running a 1792-node commodity PC Linux cluster. provides basic building blocks needed for higher-level protocols implement scalable, low-overhead communication. has several unique characteristics that differentiate it other high-performance system-area data movement layers. discusses these...
Today's high-end massively parallel processing (MPP) machines have thousands to tens of processors, with next-generation systems planned in excess one hundred thousand processors. For such scale, efficient I/O is a significant challenge that cannot be solved using traditional approaches. In particular, general purpose file limit applications standard interfaces and access policies do not scale will likely performance bottleneck for many scientific applications. this paper, we investigate the...
This paper describes our vision for Hobbes, an operating system and runtime (OS/R) framework extreme-scale systems. The Hobbes design explicitly supports application composition, which is emerging as a key approach applications to address scalability power concerns anticipated with coming architectures. We make use of virtualization technologies provide the flexibility support requirements components different node-level systems runtimes, well mappings onto hardware. describe architecture...
Infiniband is becoming an important interconnect technology in high performance computing. Recent efforts large scale deployments are raising scalability questions the HPC community. Open MPI, a new open source implementation of MPI standard targeted for production computing, provides several mechanisms to enhance scalability. Initial comparisons with MVAPICH, most widely used implementation, show similar but much better characteristics. Specifically, small message latency improved by up 10%...
This paper describes a portable benchmark suite that assesses the ability of cluster networking hardware and software to overlap MPI communication computation. The Communication Offload MPI-based Benchmark, or COMB, uses two methods characterize messages make progress concurrently, with computational processing on host processor(s). COMB measures relationship between bandwidth CPU availability.
The Program Dependence Web (PDW) is a program representation that can be directly interpreted using control-, data-, or demand-driven models of execution. A PDW combines single-assignment version the with explicit operators manage flow data values. viewed as an augmented Graph . Translation to provides basis for projects compile Fortran onto dynamic dataflow architectures and simulators. second application construction various compositional semantics dependence graphs.
Efficient data movement is an important part of any high-performance I/O system, but it especially critical for the current and next-generation massively parallel processing (MPP) systems. In this paper, we discuss how scale, architecture, organization proposed MPP systems impact design data-movement scheme system. We also describe analyze approach used by lightweight file (LWFS) project, compare that to more conventional protocols small mid-range clusters. Our results indicate strategy LWFS...
Instruction-level simulation is necessary to evaluate new architectures. However, single-node cannot predict the behavior of a parallel application on supercomputer. We present scalable simulator that couples cycle-accurate node with supercomputer network model. Our executes individual instances IBM's Mambo PowerPC hundreds cores. integrated NIC emulator into and model instead fully simulating it. This decouples simulators makes our design scalable.
Infiniband is becoming an important interconnect technology in high performance computing. Efforts large scale deployments are raising scalability questions the HPC community. Open MPI, a new open source implementation of MPI standard targeted for production computing, provides several mechanisms to enhance scalability. Initial comparisons with MVAPICH, most widely used implementation, show similar but much better characteristics. Specifically, small message latency improved by up 10%...
Large MPP systems can neither solve grand-challenge scientific problems nor enable large scale industrial and governmental simulations if they rely on extensions to workstation system software. At Sandia National Laboratories we have developed, with our vendors, a new architecture for high-end computing. Highest performance is achieved by providing applications light-weight interface collection of processing nodes. Usability provided creating node partitions specialized user access,...
Abstract In the early 1990s, researchers at Sandia National Laboratories and University of New Mexico began development customized system software for massively parallel ‘capability’ computing platforms. These lightweight kernels have proven to be essential delivering full power underlying hardware applications. This claim is underscored by success several supercomputers, including Intel Paragon, Accelerated Strategic Computing Initiative Red, Cray XT series systems, each having established...
Article Free Access Share on A massively parallel adaptive finite element method with dynamic load balancing Authors: K. D. Devine Department of Computer Science, Rensselaer Polytechnic Institute Troy, NY and Sandia National Laboratories LaboratoriesView Profile , J. E. Flaherty S. R. Wheat Massively Parallel Computing, Research Laboratory, Dept. 1424 Laboratories, Albuquerque, NM A. B. Maccabe The University New Mexico, Authors Info & Claims Supercomputing '93: Proceedings the 1993 ACM/IEEE...