IREN outlined an aggressive expansion strategy for AI infrastructure, anchored by a new $3.4 billion AI Cloud contract with NVIDIA and a broader strategic partnership focused on deploying up to 5GW of NVIDIA DSX-aligned AI infrastructure across North America, Europe, and APAC. The companies said the partnership will combine NVIDIA’s DSX AI factory architecture with IREN’s expertise in power, land acquisition, GPU deployment, data center operations, and large-scale infrastructure delivery. The announcement positions IREN among a growing class of infrastructure providers pivoting from cryptocurrency mining toward hyperscale AI compute services.
The new five-year contract with NVIDIA centers on air-cooled Blackwell GPU deployments within 60MW of IREN’s existing Childress, Texas data center campus, with ramp expected to begin in early 2027. Beyond the initial deployment, the companies said future infrastructure expansion will focus heavily on IREN’s 2GW Sweetwater campus in Texas, which is expected to become a flagship deployment for NVIDIA’s DSX architecture. NVIDIA and IREN also plan to collaborate on deployment of NVIDIA accelerated compute systems in DSX AI factories serving AI-native startups and enterprise customers. Separately, NVIDIA received a five-year right to purchase up to 30 million IREN shares at $70 per share, representing a potential $2.1 billion equity investment subject to regulatory approvals and other conditions. IREN said its AI Cloud annual recurring revenue under contract now stands at approximately $3.1 billion, with a target of reaching $3.7 billion ARR by the end of calendar 2026.
IREN also continued broadening its AI infrastructure stack through acquisitions and geographic expansion. The acquisition of Mirantis adds orchestration, software, and support capabilities intended to help operate and manage GPU infrastructure at scale. The acquisition of Nostrum adds 490MW of data center capacity in Spain along with a larger European development pipeline. Financially, the company reported Q3 FY26 revenue of $144.8 million, down from $184.7 million in Q2 FY26, reflecting lower Bitcoin mining activity and hardware decommissioning ahead of GPU deployments. Adjusted EBITDA declined to $59.5 million, while net loss widened to $(247.8) million due largely to non-cash impairment charges associated with retired mining infrastructure.
- Signed a $3.4 billion, five-year AI Cloud contract with NVIDIA for Blackwell GPU deployments
- Entered a strategic partnership with NVIDIA targeting deployment of up to 5GW of DSX-aligned AI infrastructure
- NVIDIA DSX architecture expected to anchor future AI factory deployments
- Sweetwater, Texas campus positioned as a flagship DSX deployment site
- NVIDIA received rights to purchase up to 30 million IREN shares at $70/share
- 2026 AI Cloud expansion to 480MW remains on track
- 2027 buildout to 1,210MW already under construction
- Childress Horizons 1–6 and Sweetwater 1 represent major near-term deployment sites
- Acquired Mirantis to strengthen orchestration, deployment, and customer support capabilities
- Acquired Nostrum, adding 490MW in Spain plus additional European development pipeline
- Existing AI Cloud contracts now represent approximately $3.1 billion ARR
- Targeting $3.7 billion ARR by the end of calendar 2026
- Reported Q3 FY26 revenue of $144.8 million and adjusted EBITDA of $59.5 million
- Transition from Bitcoin mining to GPU-based AI Cloud infrastructure accelerated during the quarter
“AI factories are becoming foundational infrastructure for the global economy,” said Jensen Huang. “Deploying these systems at scale requires deep integration across the full stack — compute, networking, software, power and operations. IREN brings the scale and infrastructure expertise to help accelerate the buildout of next-generation AI infrastructure globally. Together, we are building for the age of AI.”
🌐 Analysis: NVIDIA has increasingly backed a new generation of AI-focused “neocloud” infrastructure providers through direct investments, GPU allocation agreements, strategic partnerships, and warrant structures tied to large-scale deployments. Earlier this year, CoreWeave disclosed that NVIDIA held an equity stake valued at approximately $900 million following CoreWeave’s public market debut, after initially investing roughly $100 million in 2023. NVIDIA has also supported companies including Lambda, which recently expanded its AI infrastructure financing capacity with a new $1 billion syndicated credit facility, and Nebius, the AI infrastructure spinout from Yandex that has positioned itself as a sovereign AI cloud provider in Europe. In 2024, NVIDIA additionally participated in funding rounds for firms such as Crusoe Energy and Together AI as the company broadened its ecosystem beyond traditional hyperscalers.
Rather than relying exclusively on Amazon Web Services, Microsoft Azure, Google Cloud, and Oracle Cloud, NVIDIA appears to be cultivating a parallel AI infrastructure ecosystem optimized specifically for accelerated compute deployments. These neocloud operators often move faster than traditional hyperscalers in deploying Blackwell and Hopper GPU clusters, while targeting AI-native startups, sovereign AI initiatives, and enterprise training workloads. The IREN structure — including the potential $2.1 billion equity participation tied to infrastructure deployment milestones — further demonstrates how NVIDIA increasingly aligns capital investment with long-duration GPU consumption and AI factory expansion strategies.
🌐 Analysis: The NVIDIA-IREN partnership highlights the emergence of AI factory infrastructure as a new strategic layer of the compute economy. NVIDIA increasingly refers to hyperscale GPU deployments as “AI factories,” emphasizing vertically integrated infrastructure that combines accelerated compute, networking, storage, orchestration software, cooling, and massive power delivery into unified production environments for AI training and inference. The reference to NVIDIA’s DSX architecture signals alignment with the company’s broader AI factory blueprint for large-scale deployment environments optimized around Blackwell GPUs and next-generation AI networking fabrics.
IREN’s evolution also reflects a broader industry shift in which former crypto infrastructure operators are leveraging existing access to grid power, renewable energy, and large-scale land assets to compete in the AI compute market. Similar transitions are underway across the sector as GPU demand outpaces traditional hyperscale capacity buildouts. IREN’s combination of liquid-cooled infrastructure for Microsoft, air-cooled Blackwell deployments for NVIDIA, and orchestration capabilities from Mirantis positions the company as a vertically integrated AI infrastructure provider rather than simply a colocation operator.
🌐 We’re tracking the latest developments in AI infrastructure, GPU clusters, and hyperscale networking. Follow our ongoing coverage at: https://convergedigest.com/category/ai-infrastructure/
🌐 We’re launching the “Data Center Networking for AI” series on NextGenInfra.io and inviting companies building real solutions—silicon, optics, fabrics, switches, software, orchestration—to share their views on video and in our expert report. To get involved, send a note to jcarroll@convergedigest.com or info@nextgeninfra.io.