Bitcoin Miner IREN Acquires Cloud Platform Mirantis for $625M to Complete AI Infrastructure Stack
IREN Limited (NASDAQ: IREN) is acquiring Mirantis, the Kubernetes-native cloud infrastructure firm with 1,500+ enterprise customers and NVIDIA AI Cloud Ready ISV status, in a $625 million all-stock deal — completing its transformation from Bitcoin miner to full-stack AI cloud platform.
IREN Limited (NASDAQ: IREN), the Australian-founded infrastructure company that began as a Bitcoin miner before pivoting aggressively toward AI data centers, announced on May 5 that it will acquire Mirantis — the Kubernetes-native cloud infrastructure and enterprise support platform — in an all-stock deal valued at approximately $625 million at signing. Mirantis brings 1,500+ enterprise customers globally, deep Kubernetes and cloud orchestration expertise, and NVIDIA AI Cloud Ready Initiative founding ISV partner status to IREN's existing GPU data center portfolio. Mirantis will operate as a standalone subsidiary post-close, continuing to serve its existing customer base while expanding IREN's managed AI cloud offering.
From Bitcoin Mining to AI Cloud: IREN's Transformation Timeline
IREN was founded in Australia in 2019 as Iris Energy, initially focused on Bitcoin mining using renewable energy. The company listed on NASDAQ in 2021 and built data centers in Canada and Texas optimized for low-cost, low-carbon Bitcoin mining. As Bitcoin mining margins compressed during the 2022–2024 bear market, IREN began a strategic reorientation: the same data center real estate, power infrastructure, and cooling systems that run Bitcoin ASICs can be repurposed to host GPU clusters for AI inference and training workloads — with substantially higher revenue per kilowatt than mining.
By 2025, IREN had announced its first colocation partnerships with AI compute customers and began marketing its data centers as GPU-ready infrastructure. The Mirantis acquisition is the final piece of the transformation: IREN now has data center hardware (the physical infrastructure layer), GPU clusters (the compute layer), and Mirantis (the software and orchestration layer that enterprise customers use to deploy and manage workloads on top of that compute). The combination is a vertically integrated AI cloud stack competitive with smaller regional cloud providers.
Mirantis has a specific technical relevance for AI workloads. The company built its reputation on OpenStack and Kubernetes — open-source infrastructure orchestration platforms widely used in enterprise environments that run distributed computing workloads. AI training and inference are exactly this kind of workload: they require orchestrating hundreds or thousands of GPU nodes in parallel, managing job queues, and allocating compute dynamically across tenants. Mirantis has operational experience deploying these systems at enterprise scale, which reduces the time it would take IREN to build the software layer organically.
The NVIDIA AI Cloud Ready Connection
Mirantis's status as a founding Independent Software Vendor partner of the NVIDIA AI Cloud Ready Initiative is strategically significant. NVIDIA's AI Cloud Ready program certifies that an ISV's software stack is validated to run efficiently on NVIDIA GPU hardware — a stamp of approval that enterprise AI customers increasingly use as a procurement shortcut when evaluating cloud infrastructure vendors. Joining IREN with Mirantis's NVIDIA certification creates a combined offering that can credibly market to enterprises as "NVIDIA-validated AI cloud," a differentiation that commodity GPU rental providers lack.
The GPU supply dynamics of 2026 make this positioning valuable. NVIDIA H200 and Blackwell GPU clusters are still supply-constrained; enterprise customers who cannot get direct GPU allocations through NVIDIA or hyperscalers are turning to certified cloud providers. IREN's growing GPU fleet, validated by Mirantis's NVIDIA partner status, positions it as a legitimate alternative to AWS, Azure, and GCP for AI inference workloads at competitive pricing — particularly for workloads where data sovereignty or physical infrastructure control matters to the customer.
The deal also closes a customer acquisition gap. Mirantis's 1,500+ enterprise customers include large organizations that have mature Kubernetes deployments managed by Mirantis. Those customers are natural prospects for IREN's GPU cloud services: they already have a billing and support relationship with Mirantis, and the post-acquisition sales motion can offer GPU cloud access as an expansion of the existing engagement rather than a cold sale.
What to Watch
The critical execution risk is organizational integration. Mirantis is a software and professional services business with a flat, engineering-driven culture; IREN is primarily an infrastructure operations company. Maintaining Mirantis as a standalone subsidiary is the right structural choice, but the revenue synergies — cross-selling GPU cloud to Mirantis's enterprise base — require coordination that standalone structures can resist. Watch IREN's Q3 2026 earnings for any disclosure on early cross-sell traction. Watch also for IREN's power procurement strategy: AI data center economics are highly sensitive to electricity costs, and IREN's legacy in low-cost renewable energy is its margin advantage. Any deterioration in its power contracts or expansion into higher-cost power markets would erode the pricing competitiveness that makes it attractive to AI cloud customers.
Want every major AI and Web3 funding round the moment it closes? Subscribe to the BlockAI News daily brief.