Gnosis Research Center
We conduct research in High-Performance Computing (HPC) memory and storage systems, especially in the design and prototype development of scalable software systems. Our current emphasis is on a scalable system software for data management and data access and on having a better understanding of data-centric systems development.
Our research covers various aspects of high-end computing:
Memory Systems Architecture
Designing and developing compile- and run-time tools to promote efficient and high-performance computer systems.
Parallel and Distributed I/O
Performance evaluation and modeling of high performance computing systems.
Resource Management
Performance optimization, including task scheduling and fast data access for high-end computing.
Reliability and Security
Enhancing reliability and security for parallel and distributed systems.
Featured projects
ChronoLog: A High-Performance Storage Infrastructure for Activity and Log Workloads
HPC applications generate more data than storage systems can handle, and it is becoming increasingly important to store activity (log) data generated by people and applications. ChronoLog is a hierarchical, distributed log store that leverages physical time to achieve log ordering and reduce contention while utilizing storage tiers to elastically scale the log capacity.
Coeus: Accelerating Scientific Insights Using Enriched Metadata
In collaboration with Sandia and Oak Ridge National Laboratories, coeus investigate the use of an active storage system to calculate derived quantities and support complex queries on scientific data (simulation and observational) as well as optimizing data placement across the storage hierarchy, with awareness of the resource limitations, to better support the scientific discovery process.
Hermes: Extending the HDF Library to Support Intelligent I/O Buffering for Deep Memory and Storage Hierarchy System
To reduce the I/O bottleneck, complex storage hierarchies have been introduced. However, managing this complexity should not be left to application developers. Hermes is a middeware library that automatically manages buffering in heterogeneous storage environments.