Welcome to the High Perfomace Computational Services at DRI
Our HPCS team offers cost effective, responsive, adaptable, and technologically current computational service and support in a secure environment.
We operate a “small shop” approach to service and support and provide superior solutions for those seeking a reliable alternative to cloud-based services. Our technological infrastructure is powerful, reliable, scalable, and secure.
Our HPCS team is dedicated to ongoing computational performance enhancements designed to improve customer experience and satisfaction while reducing costs. Ongoing enhancements include:
- Performance tuning the Weather Research & Forecasting Model (WRF) mode, which is employed in atmospheric research and meteorological forecasting
- Compute nodes utilize Intel Knights Landing (KNL) processors and Omni-Path Architecture (OPA) networking
- Improving job pre-emption within the scheduler for job prioritization
- Creating better client reporting/graphing tools
- Cost estimation
Other virtualized, cloud-based systems (“vCPU”) share their computing resources among many clients. With DRI’s HPCS, you pay only for the actual services/resources you use versus the virtual ones you don’t; our cost structure is competitive, transparent, and straightforward. We charge our customers by the CPU hours used and bill monthly for simplicity. There are no add-on charges for storage, networking, or other processes.
Our HPCS servers, storage, and networks run on physical equipment that is securely housed at DRI and managed by on-site staff. Continuous video surveillance safeguards our HPCS facility and our state-of-the-art equipment is locked in server racks housed within an access-controlled server room. In addition, we require secure, two-factor authentication for all system users. Separated, secure networking provides complete isolation of client workloads when necessary.
- Joe Magallanes, Principal, President and CEO of Karagozian & Case (www.kcse.com)