Skip to documentation
Documentation menu

Methodology UCPI-H200-SXM-LISTED

On this page

UCPI-H200-SXM-LISTED Child Specification

Status: proposed child specification, version 0.1.0-draft. Not launched. Prepared 14 September 2026. A child of the UCPI-LISTED-GPU family specification under the Urdais Compute Price Index family; everything not stated here is inherited from those documents unchanged. A different economic object from any accessible-offer child.

Instrument

NVIDIA H200, SXM form factor, 141 GB HBM3e, full physical device, per-accelerator allocation class. The H200 NVL is a different physical product (a PCIe board pair with NVLink bridge) and is a different instrument; it is never collapsed into this one.

Admitted upstream SKUs

Price of Compute canonical SKU H200-SXM (the source separates H200-NVL). Admitted as Grade C identity.

Seller topology evidence, 14 September 2026

Verda: 1x H200 SXM5 141GB instance on its own price surface. Runpod: H200 141 GB listed per GPU with per-GPU host resources (Secure Cloud). Hyperstack: H200 SXM listed per GPU; not present in the source's H200 SXM rows on first retrieval. CoreWeave: HGX H200 sold as an eight-accelerator node, excluded WHOLE_NODE_REQUIRED. Nebius: excluded SELLER_LEGAL_IDENTITY_UNRESOLVED (see the H100 child). Vast.ai: platform aggregate, excluded.

Version History

0.1.0-draft, 14 September 2026: initial child. No production effective date.