On Jan. 2, 2025, GWireless retired. Join eduroam, GW’s primary wifi network. For the best experience, forget the GWireless network and then set up your eduroam connection on your devices.
Advancing Research at GW by expanding HPC
GW IT, in collaboration with various academic and administrative units, is updating and expanding its flagship High-Performance Computing (HPC) platform, Pegasus. This initiative aims to enhance research at GW by expanding HPC infrastructure to improve reliability and provide increased computing capabilities.
GW is a top-100 research university with a diverse scientific portfolio of over $350M annual research expenditures (2023). GW’s continued competitiveness in computational and data intensive research is a major contribution to GW research efforts.
The expansion of the Pegasus HPC platform will implement state-of-the-art GPU-based systems, including the Nvidia Grace Hopper “superchip” and L40S GPU. This advanced infrastructure enables groundbreaking work across various fields, facilitating new innovations and discoveries.
Key Benefits for Researchers
- 10x higher performance compared to the previous generation, powered by the Nvidia Grace Hopper superchip, which combines an Nvidia Grace CPU with a Hopper GPU. The addition of the L40S GPU further increases the processing capacity for demanding computational and visualization tasks. This is crucial for accelerating research in areas such as advanced deep learning, AI research, and computational fluid dynamics.
- High-speed, low-latency components for networking, memory, storage, and file systems, further optimizing the computing power. Increased memory capabilities allow researchers to handle larger datasets, which is crucial for fields like genomics, climate modeling, and big data analytics.
- Improved performance, enabling faster processing of complex simulations and computations. The memory bandwidth of the Grace Hopper Superchip, combined with the AI and graphics capabilities of the L40S GPU, allow multiple high-memory applications to run simultaneously without performance degradation.
Image
Questions & Feedback
If you have questions or feedback about this change, please share them with us by submitting a Research Technology Services Request, and a member of the project team will be in contact with you.
Classroom Support
202-994-7900
Classroom Search
Phone
202-994-4948
24 hours / 7 days a week
Walk-In
Walk-In Support Centers
Knowledge Base
Explore our knowledge base for how-to articles and guides.