Back to stories
Industry

NVIDIA and IREN Announce 5GW AI Infrastructure Partnership Backed by $2.1B Warrant

Michael Ouroumis2 min read
NVIDIA and IREN Announce 5GW AI Infrastructure Partnership Backed by $2.1B Warrant

NVIDIA and IREN unveiled one of the largest single-vendor AI infrastructure partnerships announced to date on May 7, with the two companies agreeing to support deployment of up to 5 gigawatts of NVIDIA DSX-aligned AI factories across IREN's global data center footprint. The deal pairs the buildout commitment with a five-year warrant giving NVIDIA the right to purchase up to 30 million IREN shares at $70 each — a potential investment of up to $2.1 billion subject to regulatory conditions. IREN shares rallied sharply after the news crossed.

A capacity commitment that dwarfs most hyperscaler campuses

Five gigawatts is roughly the scale of an entire mid-sized country's data center fleet, and represents NVIDIA's most explicit move yet to lock in physical capacity ahead of demand for its next-generation systems. The companies said the rollout will focus initially on IREN's 2-gigawatt Sweetwater campus in Texas, which they expect to serve as a flagship site for NVIDIA's DSX AI factory architecture. The remaining capacity is described as being available across IREN's broader pipeline over time, without firm dates.

"AI factories are becoming foundational infrastructure for the global economy," NVIDIA founder and CEO Jensen Huang said in the announcement. "Deploying these systems at scale requires deep integration across the full stack."

Why the warrant structure matters

The $2.1 billion warrant is unusual in scale and structure. Rather than a direct equity investment or a prepayment for compute, it gives NVIDIA optionality to take a meaningful stake in IREN over five years at a fixed strike price, aligning the chip vendor's upside with IREN's execution on the buildout. It also echoes the financing-and-equity hybrid models that have characterized recent multibillion-dollar tie-ups across the AI infrastructure stack, where capital, supply commitments, and customer relationships are increasingly bundled into a single agreement.

IREN co-founder and co-CEO Daniel Roberts framed the rationale around capability stacking. "This partnership combines NVIDIA's AI systems and architecture leadership with IREN's expertise across power, land, data centers, GPU deployment and infrastructure operations," he said.

IREN's positioning

IREN operates as a vertically integrated AI cloud provider, offering large-scale data centers and GPU clusters for training and inference. The company leverages a portfolio of grid-connected land and power assets in renewable-rich regions across North America, Europe, and APAC — the kind of pre-secured power capacity that has become the binding constraint for hyperscale AI deployments in 2026.

Implications

For NVIDIA, the deal is another lever to ensure DSX-class systems land in operational data centers as fast as power and shells can be brought online. For IREN, it transforms a former bitcoin-mining-rooted infrastructure operator into a credible top-tier AI cloud counterparty, with NVIDIA's brand, architecture roadmap, and potential equity participation behind it. For the broader market, the agreement underscores how AI capacity planning has shifted from quarterly chip allocations to multi-year, gigawatt-scale commitments tied to specific physical sites — and how the line between supplier and shareholder continues to blur.

Learn AI for Free — FreeAcademy.ai

Take "AI for Business: Practical Implementation" — a free course with certificate to master the skills behind this story.

More in Industry

Anthropic Signs $1.8 Billion Cloud Deal With Akamai, Largest in CDN's History
Industry

Anthropic Signs $1.8 Billion Cloud Deal With Akamai, Largest in CDN's History

Anthropic locked in a seven-year, $1.8 billion cloud infrastructure commitment with Akamai, sending the legacy CDN provider's shares up roughly 26% on what CEO Tom Leighton called the largest customer deal in company history.

8 hours ago3 min read
Cloudflare Cuts 1,100 Jobs in 'Agentic AI-First' Pivot Despite Q1 Beat
Industry

Cloudflare Cuts 1,100 Jobs in 'Agentic AI-First' Pivot Despite Q1 Beat

Cloudflare is laying off about 1,100 employees — 20% of its workforce — to restructure around what CEO Matthew Prince calls an 'agentic AI-first operating model,' even as Q1 revenue grew 34% to $639.8 million.

12 hours ago2 min read
OpenAI, Nvidia, Microsoft, AMD, Intel and Broadcom Unite Behind MRC Networking Protocol for Gigascale AI
Industry

OpenAI, Nvidia, Microsoft, AMD, Intel and Broadcom Unite Behind MRC Networking Protocol for Gigascale AI

OpenAI announced Multipath Reliable Connection (MRC), a new RDMA-over-Ethernet protocol co-developed with Nvidia, Microsoft, AMD, Intel and Broadcom. The protocol is already running on OpenAI's largest GB200 clusters and has been contributed to the Open Compute Project.

1 day ago2 min read