Department of Energy Contracts IAI to Develop an Intelligent Security Assessment Tool for HPC Systems via Automated Learning and Data Analytics

It is important to have enhanced cybersecurity on High Performance Computing (HPC) systems that enable large-scale information processing for critical national infrastructure, government, and commercial enterprise systems. Software protection and associated tools are a first line of defense for run-time integrity in a HPC system, but these cannot be easily extended to provide security solutions beyond the software level. Major challenges to building a system-wide security analysis tool for assessing the entire HPC system’s security include providing efficient data processing, customized feature extraction, and balance between security needs and overhead. IAI and collaborators at the New Jersey Institute of Technology and University of South Florida will develop a data analytics based system-wide security assessment tool for HPC systems to ensure system-wide security monitoring and assessment. The team will leverage existing or future HPC profiling and tracing tools to provide real-time data collection. This information, combined with HPC user or program behavioral knowledge, system logs and network traffic data, will be used for cyber security assessment. This innovative method uses passive data collection, and leverages existing HPC profiling and tracing tools, thus requiring a minimum of extra monitoring tools to be installed on HPC systems. The team will design the security assessment system architecture and workflow, design and implement initial development of core functional modules, and build a proof-of concept prototype for demonstration and performance evaluation. These techniques, tools, and software will enhance cyber security for HPC systems, and can be applied to a broad range of HPC centers and Large-Scale Distributed Computer Systems in government agencies as well as industry and university settings.