Cluster · AWS · ultraserver
AWS Trainium 2 UltraServer
AWS's answer to NVL72: 64 Trainium 2 chips lashed together with NeuronLink. Anthropic's Project Rainier compute unit.
Key facts
- Chips per instance
- 64
- Interconnect
- NeuronLink (2 TB/s aggregate per UltraServer)
- Cooling
- water
- Announced
- 2024-12-03
- GA / release
- 2025-01-01
Chips inside
Where you'll find it
Data centers publicly running the chips inside this design. Per-rack counts are rarely disclosed, so these are presence rows: this campus runs these chips.
- Amazon 'Bridgefield' GWU Ashburn Campus · AWS Trainium 2 capacity across the Ashburn / GWU Bridgefield campus
- Amazon Project Rainier Data Center (New Carlisle) · Anthropic's compute (Project Rainier): hundreds of thousands of Trainium 2 chips
- AWS Susquehanna (Cumulus) Nuclear-Coupled Data Center Campus · Anthropic Project Rainier training cluster — Trainium 2 at scale (960 MW envelope)
Vendor page: https://aws.amazon.com/ai/machine-learning/trainium/
See also: all rack designs · chip catalog · AWS.