Test report DSG-4819 · Rev B · tested November 16, 2026
AI Datacenter InfrastructureDevice under test
OpenAI signs 457MW AI capacity deal with TensorWave
OpenAI’s six-year, 457MW TensorWave deal uses AMD Instinct MI455X Helios racks, with operations due in the first half of 2027 and full deployment by early 2028.
- Read
- 3 min
- Words
- 558
- Node
- 7nm
- Operator
- Amara Osei
Spec summary
- OpenAI signed a six-year contract for 457MW of TensorWave capacity.
- Operations are expected in the first half of 2027, with full deployment by early 2028.
- One Helios rack provides up to 2.9 exaflops FP4, 1.4 exaflops FP8, 31TB of HBM4 and 1.7PB/s of memory bandwidth.
- AMD claims 15% more peak FP4, 50% more HBM capacity, 6% more HBM bandwidth and 50% more scale-out bandwidth than Nvidia’s Vera Rubin NVL72.
- TensorWave raised $350 million in June 2026 and claims 2GW of long-term US data center capacity.

By Amara Osei
AI lab OpenAI has signed a six-year agreement with TensorWave for 457MW of dedicated IT capacity, the AMD-focused neocloud’s largest deployment contract to date.
The agreement covers AMD Instinct MI455X-powered Helios rack-scale systems across multiple data center locations. TensorWave will operate the systems as dedicated, single-tenant capacity for OpenAI. Operations are expected to begin in the first half of 2027, with full deployment by early 2028.
What hardware sits inside the 457MW deployment?
AMD launched the Helios racks in July. Their rack-scale architecture combines AMD Instinct MI455X GPUs, 6th-generation AMD EPYC “Venice” server CPUs, AMD ROCm software and AMD Pensando networking in one system.
A single rack provides up to:
- 2.9 exaflops of peak FP4 performance
- 1.4 exaflops of peak FP8 performance
- 31 terabytes of HBM4 memory
- 1.7 petabytes per second of memory bandwidth
In practical terms, the 457MW deployment uses a rack that integrates accelerators, processors, software and networking.
How does AMD position Helios against Nvidia?
Against published specifications for Nvidia’s Vera Rubin NVL72 rack, AMD claims Helios delivers:
- 15% more peak FP4 performance
- 50% more high-bandwidth memory capacity
- 6% more HBM bandwidth
- 50% more scale-out bandwidth
AMD also claims it will deliver up to 30% more tokens per dollar.
On AMD’s figures, memory capacity and scale-out bandwidth show the largest claimed gains, both at 50%.
What do the executives say about delivery?
Piotr Tomasik, president and co-founder of TensorWave, said:
“AI infrastructure is entering a new phase, where the ability to deliver hundreds of megawatts of advanced compute reliably and at speed will separate the companies that can support frontier AI from those that cannot.”
He added:
“TensorWave was built to prove that AMD infrastructure could operate at the scale and reliability that the world’s most demanding AI companies require. This deployment takes that work to an entirely new level, with TensorWave helping deliver a meaningful portion of the infrastructure behind OpenAI and AMD’s next generation of AI compute.”
Uday Ruddarraju, CTO for compute at OpenAI, said:
“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. We’re excited to advance our previously announced partnership with AMD through this agreement with TensorWave, which will help us continue scaling AI to serve more people and organizations around the world.”
Both executives linked the contract to the delivery of reliable compute at frontier-AI scale.
What supports TensorWave’s expansion?
TensorWave secured $350 million in a Series B funding round in June 2026, which the company said would enable it to expand its AI infrastructure footprint. TensorWave claims to have secured 2GW of long-term data center capacity across the United States.
TensorWave is also a known customer of TecFusions. In January 2026, it signed an additional 20MW capacity agreement with TecFusions for data centers in Pennsylvania and Arizona.
TecFusions had previously been set to lease capacity at Fermi’s Project Matador, but the companies ended that arrangement earlier this month.
TensorWave enters the OpenAI deployment with $350 million in new funding and a claimed 2GW US capacity footprint.
The delivery schedule is clear: operations start in the first half of 2027 and reach full deployment by early 2028.
via datacenterdynamics.com (Source)
More from Amara Osei
Show full bio
Staff writer covering money, funding and market shifts around hardware layer of artificial intelligence at Die Signal.
698 articles
Same lot · LOT-362C
- DSG-671165nmMicrosoft, AMD Unveil Helios Rack for Azure AI Inference
- DSG-680245nmOracle Orders 50,000 AMD MI450 GPUs in $3.5–4B Helios Deployment
- DSG-425428nmAMD Helion: 32,768-Core CPU-Only Rack Targets Agentic AI Workloads
- DSG-468765nmAMD's Instinct MI455X Pairs 432 GB HBM4 with 72-GPU Helios Rack
- DSG-29737nmAMD's 97% Data Center Jump vs Nvidia's 30x Claim: Where AI Buyers Stand