Exclusive video content at insideHPC

In this video, Dr. Rupak Biswas from NASA presents: NASA Advanced Computing Environment for Science and Engineering. Recorded at the National HPCC conference 2012 in Newport.

In addition to providing production supercomputing cycles to NASA scientists and engineers, HECC is also responsible as the Agency’s expert in evaluating emerging supercomputing technologies and maturing the most appropriate ones into the production environment. For instance, HECC has 200 of the latest GPUs from Nvidia, several of which are integrated directly into Pleiades. GPUs can be an extremely efficient way to process massive amounts of observational data. But other accelerator technologies are being critically evaluated as well. One interesting option is Intel’s Many Integrated Core (MIC) architecture that offers a critical advantage over GPUs in that code conversions to CUDA or OpenCL are not required to run applications – only compiler directives. However, full utilization of MIC may require code modifications. HECC is also investigating other advanced IT technologies such as cloud computing for science and engineering applications, collaborative environments such as the NASA Earth Exchange (NEX), special-purpose hardware/software solutions for Big Data analysis, and quantum computing for difficult non-polynomial time optimization problems. The overall goal is for HECC to provide a consolidated bleeding-edge environment for all of NASA’s computational and analysis requirements.

Note: audio issues with this program have been repaired.

In this video, Alex Kent from Los Alamos National Lab presents: Cyber Security Defense Using HPC.

Abstract:
Adversaries in the cyber domain continue to escalate their use of more sophisticated attacks and associated detection countermeasures. As a result, the difficulty and complexity of finding such adversaries and their attacks increasingly challenges cyber defenders. Traditionally, high performance and parallel computing (HPC) has been a successful tool in tackling complex problems, particularly over large data sets. Yet it has only been recently that HPC has successfully impacted the cyber defense problem. This talk will introduce the cyber defense problem space and its escalating nature. Three key areas where HPC is making contemporary and tangible improvements to cyber defense will then be discussed: distributed data storage and query, streaming data analysis, and large-scale graph analysis. Each of these areas will include overviews of both research and operational systems currently at the Department of Energy national laboratories that are leveraging core HPC capabilities for improved cyber defense. In addition, the importance and use of both message passing (e.g. MPI) and map/reduce (e.g. Hadoop) approaches to parallel processing will be highlighted. The talk will conclude with a discussion of the critical importance of HPC in forward-looking cyber security defense and where future directions may lead.”

Recorded at the National HPCC conference 2012 in Newport.