egov.mn
Монголоор унших

TensorWave signs six-year deal to supply OpenAI with 457MW of AMD-powered AI capacity

The AMD-backed neocloud will provide OpenAI with dedicated capacity on AMD Instinct MI455X GPUs across multiple US data center sites, with first capacity due in the first half of 2027.

This story was prepared automatically by software.

TensorWave company logo
TensorWave company logo

TensorWave, an AI cloud provider backed by AMD Ventures, announced on November 16 a six-year agreement with OpenAI to provide 457 megawatts of IT capacity across multiple US data center sites, powered by AMD Instinct MI455X GPUs.

The infrastructure will be purpose-built for AMD's Helios rack-scale systems and operated as dedicated, single-tenant capacity for OpenAI. Capacity is expected to begin coming online in the first half of 2027, with full deployment expected by early 2028. Neither company disclosed financial terms or named the sites.

Part of the AMD-OpenAI buildout

TensorWave said it is the first neocloud publicly announced as an infrastructure provider supporting AMD and OpenAI's previously announced strategic partnership. AMD and OpenAI announced a deal in October 2025 to deploy 6 gigawatts of AMD GPUs. A TensorWave executive told the Las Vegas Review-Journal the new agreement represents around 8% of that deal, and that it will grow TensorWave's capacity from 35 MW to more than 450 MW and its workforce from 225 to more than 1,000 by the end of next year.

"As demand for AI continues to grow, expanding access to reliable, large-scale compute is critical to our ability to build and deliver increasingly capable models," said Uday Ruddarraju, OpenAI's chief technology officer for compute. Vamsi Bopanna, AMD's senior vice president for AI, called TensorWave a strong partner in scaling AMD Instinct infrastructure.

According to DatacenterDynamics, Helios combines MI455X GPUs, AMD Epyc "Venice" CPUs, ROCm software and Pensando networking, and AMD claims 15% more peak FP4 performance and 50% more high-bandwidth memory capacity than Nvidia's Vera Rubin NVL72 rack.

Sources: Business Wire / TensorWave, DatacenterDynamics, Las Vegas Review-Journal

Share
Next storyUS and nine partners sign San Francisco Declaration to safeguard AI supply chains
Newsletter

Get the latest news delivered to your inbox.

Related articles