Case Study: Georgia Tech & Penguin Computing
At the Georgia Institute of Technology, Penguin Computing built the Myriad cluster, a system with over 10,000 CPU cores and a 100 TFLOP (teraflop) theoretical performance maximum. Deployed at the Center for the Studies of System Biology headed by Dr. Jeffrey Skolnick, Myriad is used for large-scale computer simulations of proteins and cell models, with the aim of accelerating the process of drug discovery, as well as the diagnosing and treating disorders such as cancer
We replaced our legacy IBM system with a TOP100 cluster from Penguin in a ‘rolling’ update without significant downtime. A daunting task that Penguin was very successful at completing. We are very happy.
Professor Dr. Jeffrey Skolnick, Director for the Center for the Study of Systems Biology, Georgia Institute of Technology
An HPC infrastructure in an academic institution needs to integrate heterogeneous compute resources for a wide variety of applications across multiple disciplines and a very diverse user base. The issues researchers face in these environments are:
- fair distribution of computational resources
- integration of resources from desktop PCs to large-scale clusters
- sizing HPC clusters for workloads that vary widely over time
- constrained budgets
- deployment timelines dictated by grant cycles
- management of local HPC resources by researchers with little HPC expertise
HPC Solutions for Academia from Penguin Computing
Penguin Computing has successfully delivered HPC solutions to hundreds of academic institutions such as University of Delaware, California Institute of Technology and University of Memphis.
In addition to hardware and software offerings Penguin provides easy and affordable on-demand access to high-end HPC resources for academic institutions. In cooperation with Indiana University Penguin’s academic cloud POD is accessible via Internet II, a high-speed national research network. The academic cloud is accessible by researchers at US institutions of higher education and federally-funded research centers.
When we had to make a decision on the server platform for our upcoming cluster deployment we looked at a number of solution providers. Given the success we have had with Penguin in the past it made the most sense to select Penguin for this deployment as well.
Daniel J. Grim, Chief Technology Officer,
Information Technologies at the University of Delaware