Empire AI

Overview

Empire AI is a New York State initiative launched by Governor Kathy Hochul in April 2024. It is a collaboration of New York State's top public and private universities. The mission of Empire AI is to equip New York's leading research institutions with high-performance computing resources for responsible artificial intelligence research and development.

In addition to RPI, other members are Columbia University, Cornell University, CUNY, SUNY, NYU, University of Rochester, RIT, Flatiron Institute, and the Icahn School of Medicine at Mount Sinai.

Institutional Allocation & Service Units

All Empire AI computer resources are assigned a value in Service Units (SUs). Each SU will equate to $0.50.

  • 1 Alpha GPU hour = 1 SU
  • 1 Beta GPU hour = 2 SU
  • 1 Grace-Grace node hour = 0.5 SU
  • 1 TB.month = 8.333 SU up to 100 TB, data over 100 TB will be free but allocated and managed separately
  • Priority queue access 2x the above
  • Shared resource access 0.5x the above
  • These above conversion factors are termed multipliers
  • Cost of a job in SUs = duration * amount of resources * multiplier

Hardware specifications

Alpha System

24 HGX Nodes

  • Nodes alphagpu01-alphagpu18
    • 8 H100 80 GB GPUs per node
  • Nodes alphagpu19-alphagpu24
    • 8 H200 80 GB GPUs per node
  • 10-400 Gb/s ConnectX-7 NIC Cards (8 for IB and 2 for Ethernet)
  • 30 TB NVMe caching space
  • 2 TB of system memory
  • 23 of the nodes serve as the main computational platform. The 24th node is primarily to serve development, testing, etc.

Non-blocking NDR fabric cabled for rail configuration

  • 8 network switches, 96 optical connections

4 service nodes

  • 2 login nodes and 2 cluster management nodes (NVIDIA Base Command with licenses for all gear)

2 data transfer nodes

2 PB of DDN Storage

  • 4 x 720 TB Flash storage (training data, snap shots)

2 PB of VAST Storage

  • home directories

Beta System (Expected Late April 2026)

 

Back to top