GPU Exchange is sourcing a complete H100 SXM training cluster — compute, fabric and storage — on behalf of a well-capitalized AI company. Turnkey and partially integrated offers are both considered.
Required Configuration
- Compute: 128× 8-GPU HGX H100 80GB SXM5 servers (1,024 GPUs total)
- Fabric: Non-blocking NDR 400Gb/s InfiniBand compute fabric, rail-optimized
- Storage: High-performance parallel storage, 2PB usable or greater
- Racks: Fully racked, cabled and labeled preferred
- Condition: New or recently decommissioned with complete service history
- Location: Specify current inventory location; in-place deployments considered
- Availability: Ready to ship, or provide a confirmed lead time
Supplier Information Requested
Please include the following with your quote:
- Complete bill of materials for compute, fabric and storage
- Rack elevations and power requirement per rack
- Deployment history and decommissioning date, if used
- Management software and licenses included
- Quantity available
- Unit price and price for the full available quantity
- Condition and cosmetic grade
- Testing status and available test reports or diagnostics
- Warranty or return rights
- Inventory location and shipping terms
- Earliest ship date
- Serial number documentation available for verification
- Any export, end-user, geographic, or resale restrictions
Commercial Requirements
- Inventory must be physically verifiable.
- Seller must demonstrate lawful ownership and authority to sell.
- Equipment must be free of liens and third-party claims.
- Any final transaction is subject to technical verification, supplier diligence, sanctions screening, export-control review, and mutually acceptable purchase documentation.
- Do not send customer-identifying or end-user-confidential information unless specifically requested.
Acceptable Responses
Identify any deviation from the requested specification clearly. The following may be quoted separately:
- H200 SXM clusters of comparable size
- Clusters of 512 GPUs or more that can be combined to reach the target
Submit To
GPU Exchange
sales@gpuexchange.net
www.gpuexchange.net