Project Status: In Progress

Project Manager:  Randy Gadell,

Project Goals and Objectifications

To build central IT infrastructure and services that allow the WashU Researcher to focus on their research instead of acquisition and support of their own IT infrastructure. This multi-year project will design, implement, and support a high-performance computing platform for WashU researchers.

Project Scope
  • Develop, design, implement, and support the advanced IT compute infrastructure in a platform-as-a-service model
  • Implement high-performance processor-based (HPC) compute services with 5,000 Intel cores
  • Develop high-performance GPU (cuda core) compute services with 500,000 NVidia GPUs
  • Implement 300TB of high-performance solid-state storage for fast program execution (scratch space)
  • Implement a 300Gb compute to storage network
  • Integrate storage services and compute services and research web services
MilestonesCompletion Date
Start Research Compute Services Project (Funded)December 2018
Develop design and statement of workMarch 2019
Issue RFP to vendorsApril 2019
Select system and vendorMay 2019
Issue orders and contract approvalsJune 2019
DeliveriesJune – August 2019
InstallationAugust – September 2019
Testing and PilotsOctober – December 2019
Early AdoptersJanuary 2020
Phase2 StartMarch 2020
Phase 2 Purchases
-Additional CPU Servers
-Two NVidia DGX A100 GPU Servers
-Network, Storage, Software Enhancements
July-August 2020
Training for Principle Investigators and StaffOn-going
General AvailabilitySeptember 2020
Core Team

Core Team

Team MemberTitleRole
James Fitzpatrick, Ph.D.Professor of NeuroscienceChair of Infrastructure and Applications Subcommittee
Randy GadellProject ManagerProgram / Project Manager
Craig PohlSenior Director, RISTechnical Lead