Advanced Computing, Mathematics and Data
Research Highlights
April 2008
Multithreaded Architectures for Emerging Applications Show Promise
Researchers at the Pacific Northwest National Laboratory (PNNL) are creating new applications and mapping existing applications to the Cray XMT platform with promising results. These promising results reinforce the Cray XMT potential to substantially accelerate data analysis and predictive analytics to address many complex challenges in energy, national security and fundamental science that traditional computing cannot do.
The Cray XMT, with its unique "massively multithreaded" architecture and large global memory, can be used to successfully execute applications that require access to terabytes of data arranged in a random and unpredictable manner. These applications, such as data discovery, bioinformatics and power grid analysis, are difficult to map to current distributed memory systems where each processor has an independent memory and to achieve scalable—proportional to the number of processors used—performance on such systems.
Researchers have obtained promising results in applying the Cray XMT to the cyber security domain involving large sets of network traffic data. Analysis was performed to detect anomalies in packet headers (snippets of Internet communication between computers) in order to locate and characterize network attacks, and to help predict and mitigate future attacks.
Promising results also were obtained in applying the Cray XMT to the biological domain. The biological domain involved solvers for Boolean Satisfiability problems with applications to biological network analysis. The work is an effort to determine if variables may be assigned in such a way as to yield "true." For example, the solution to Boolean equations can determine whether electronic components will function according to their design or not.
"Our preliminary results indicate that these challenging applications can achieve scalable parallelism on the Cray XMT platform beyond what is possible on mainstream high-performance computing platforms," said PNNL scientist Daniel Chavarria. "Faster solutions to these problems can lead to earlier detection of cybersecurity threats as well as enhanced and more precise understanding of biological properties."
A paper discussing the preliminary results of the two applications will appear in the 2008 summer issue of SciDAC Review.
Acknowledgments
Sponsors - Pacific Northwest National Laboratory's Data Intensive Computing Initiative; DOE's Office of Advanced Scientific Computing Research.
Research Team - Dr. Daniel Chavarria-Miranda, Dr. Chad Scherrer and Dr. Heidi Sofia (PNNL); Kristyn Maschhoff, Cray, Inc.
