Microsoft is already running AI workloads on IREN's infrastructure. Nvidia has now validated another layer of its platform. But according to co-CEO Daniel Roberts, those milestones point to a bigger story: in the AI infrastructure race, owning the entire stack—not just the chips—could become the industry's defining advantage.
Speaking on IREN's fiscal 2026 earnings call, Roberts argued that the company's strategy extends far beyond acquiring Nvidia GPUs. Instead, he said IREN is building a vertically integrated AI infrastructure platform spanning power, data centers, compute, and software—an approach he believes creates more value than relying on third-party providers for critical pieces of the stack.
Microsoft and Nvidia Reinforce IREN's AI Stack Strategy
IREN highlighted a series of milestones that underscore that strategy.
The company said Horizon 1—the first of four planned 50-megawatt AI cloud deployments under its Microsoft contract—was delivered during the quarter. Roberts also noted that the company's 2026 contracted annual recurring revenue has reached $4 billion, with $1 billion already operational, while another $700 million of ARR tied to its Nvidia Cloud contract is expected to ramp in 2027.
On the software side, Roberts pointed to another milestone announced alongside the earnings release.
“Just today, Mirantis was named an inaugural Nvidia Certified Hypervisor. So we've now got Nvidia validation at the software layer as well as the hardware.”
Mirantis, which IREN recently acquired, enables the company to offer managed AI cloud services alongside bare-metal GPU infrastructure, expanding the range of customers it can serve.
IREN CEO Says Owning Every Layer Creates More Value
For Roberts, the Microsoft deployment and Nvidia software certification illustrate a broader investment thesis rather than isolated achievements.
“Why own all three?” he asked, referring to infrastructure, GPUs, and software.
“Because each layer makes the one underneath it worth more. A grid connection is worth more with a data center on it. Worth more again with GPUs inside. More again with services wrapped around the customer.”
He contrasted IREN's approach with much of the broader AI infrastructure market.
“Most of this market rents at least one of those layers; we own the entire stack.”
Management argued that vertical integration gives IREN greater flexibility to serve different customer types, from hyperscalers seeking bare-metal GPU clusters to enterprise AI developers that require managed cloud services and orchestration software.
The Mirantis acquisition also opens the door to new offerings such as on-demand AI compute and managed services, complementing the company's long-term cloud infrastructure contracts.
What Investors Should Watch
The AI infrastructure story has largely centered on access to Nvidia's latest chips. IREN's latest earnings call suggests management believes the competitive landscape is shifting toward a broader question: who owns the infrastructure surrounding those GPUs.
That thesis is now supported by tangible milestones. Microsoft has begun taking delivery under its AI cloud agreement, Nvidia has validated IREN's software layer through Mirantis, and management says each additional layer strengthens the economics of the one below it.
For investors, the next test will be whether that vertically integrated model translates into sustained customer growth, expanding margins, and higher-value AI cloud contracts as demand continues to scale.
