Resources

1. Experimental Computing Laboratory (ExCL)
2. Future Technologies Colloquium Series
3. Publications
4. Call for papers
5. Visitor Information page (maps, lodging information, restaurants)
6. Jobs
7. Software
8. FT Home Page on CSM website

Staff

1. David Bader (joint with Georgia Institute of Technology)
2. Anthony Danalis (joint with University of Tennessee-Knoxville)
3. Deborah Holder (Group Secretary)
4. Gabriel Marin
5. Collin McCurdy
6. Jeremy Meredith
7. Kenneth Roche
8. Philip Roth
9. Thomas Sterling (joint with Louisiana State University)
10. Olaf Storaasli
11. Vinod Tipparaju
12. Jeffrey Vetter (Group Leader)
13. Weikuan Yu (joint with Auburn University)

Visitors

1. Ernest Cochran
2. Karthik Babu Parasurama Vijayakumar, North Carolina State University
3. Ruoke Yang, Oak Ridge High School

Research

1. Early Evaluation of HPC Systems
2. Emerging Architectures
3. Parallel IO
4. Productive Programming Environments
5. Performance Prediction and Analysis
6. Visualization

Sponsors

News

2009-05-23  --  ORNL and Sun Researchers Demonstrate Parallel NFS Over Lustre at IPDPS 2009.  Researchers from the Future Technologies group at ORNL (Jeffrey Vetter and Weikuan Yu) and at Sun Microsystems (Oleg Drokin) have designed and implemented the first prototype of Parallel NFS over the scalable Lustre file system. Parallel NFS (pNFS) is an emergent open standard for parallelizing data transfer over a variety of I/O protocols. In a paper presented at the 23rd IEEE International Parallel and Distributed Processing Symposium, the team presented the design, implementation, and evaluation of lpNFS, a Lustre-based Parallel NFS. The benefits of using pNFS include portability across a range of back-end file systems with no changes to the client-side operating system. The initial performance evaluation shows that the performance of pNFS is comparable to that of native Lustre under many I/O workloads. Given these results, lpNFS appears to be a promising approach to providing a scalable, high performance, portable, datacenter-wide file system for DOE computing facilities.  

2009-05-23  --  ORNL Researchers accelerate materials application with Graphics Processing Units (GPUs).  In a recent article (URL: http://dx.doi.org/10.1016/j.parco.2008.12.004) in the journal ‘Parallel Computing,’ a team of ORNL researchers (Jeremy S. Meredith, Gonzalo Alvareza, Thomas A. Maier, Thomas C. Schulthess, Jeffrey S. Vetter) show how they have accelerated the Quantum Monte Carlo simulation code, named DCA++, using graphics processing units (GPUs) as general-purpose computational devices (also known as GP-GPUs). While initially designed for real time rendering, the high performance and relatively low cost makes GPUs a desirable target for scientific computation. Recent efforts in the community have been addressing the programming challenges, with new languages such as CUDA and OpenCL being widely adopted. However, the original task of GPUs - rendering - has traditionally kept accuracy as a secondary goal, and sacrifices have sometimes been made as a result. In fact, much deployed GPU hardware is only capable of single precision arithmetic, and even this accuracy is not always equivalent to that of a commodity CPU. In this paper, the team investigated the accuracy and performance characteristics of GPUs on DCA++, including results from a preproduction double precision-capable GPU. They then accelerated the full DCA++ application, while concurrently investigating its tolerance to the different levels of arithmetic precision available in GPUs. The results show that while DCA++ has some sensitivity to the arithmetic precision, the single-precision GPU results were comparable to single-precision CPU results. Acceleration of the code on a fully GPU-enabled cluster showed that any remaining inaccuracy in GPU precision was negligible. Sufficient accuracy was retained for scientifically meaningful results while still showing significant speedups; the full parallel runtimes on the GPU cluster were five times faster than that on commodity microprocessors alone.  

2008-11-20  --  Meredith wins Gordon Bell Prize.  Jeremy Meredith was part of an ORNL team that won the Gordon Bell Prize at SC08 in Austin, Texas. The team led by Thomas Schulthess of ORNL received the prestigious 2008 Association for Computing Machinery (ACM) Gordon Bell Prize after attaining the fastest performance ever in a scientific supercomputing application: 1.3 petaflops on DCA++, an application that simulates superconducting materials. Meredith's investigation of porting DCA++ to Graphics Processors, like NVIDIA's Tesla, triggered an investigation of the application's requirements for double-precision arithmetic. After careful analysis, the team realized that they could retain accuracy while using single-precision arithmetic for a portion of their calculations. Since single-precision computation is typically faster on today's architectures, the team was able to realize a significant performance improvement from this change. 

2008-11-15  --  Future Tech researchers contribute to SC08 conference.  This week in Austin, Texas, researchers from ORNL’s Future Technologies Group are contributing to multiple events at SC08, the premier international conference for high-performance computing (HPC), networking, storage and analysis. Members of the FT group have contributed to four technical papers, one Gordon Bell Finalist submission, one poster, and multiple birds-of-a-feather sessions. The four technical papers are titled “Early Evaluation of BlueGene/P”, “Wide-Area Performance Profiling of 10GigE and Infiniband Technologies”, “High Performance Multivariate Visual Data Exploration for Extremely Large Data”, and a “New algorithm to enable 400+ TFlop/s sustained performance in simulations of disorder effects in high-Tc superconductors,” which is also a Gordon Bell Finalist. The poster is titled “Modeling Assertions for Petascale Applications and Systems.” Members of the group were also involved in numerous conference committees: technical papers, tutorials, and disruptive technologies. For more information and schedules, visit sc08.supercomputing.org, scyourway.nacse.org, and our ORNL SC08 page.  

2008-06-02  --  Interns from around the country to work with ORNL's Future Technologies Group this summer.  Joining us this summer are: Tristan Cuevas from Kent State University, Ashley DeFlumere from Mount Holyoke College, Andrew Gearhart from University of Delaware, Charles Lively from Texas A&M, Kelly Livingston from University of Delaware, Kyle Spafford from Georgia Institute of Technology, and Nirmal Thacker from Georgia Institute of Technology. They will examine a number of technical research topics that include emerging architectures and HPC system software.  

--- Read more articles at our news archive. ---