IREN stock soars on $2.8 billion in new AI cloud contracts
IREN Limited (NASDAQ: IREN), a key AI cloud infrastructure provider, announced on Monday, July 20, 2026, it has secured $2.8 billion in new multi-year AI cloud services contracts. This significant influx of deals with leading AI developers sent the company’s stock surging by as much as 16%.
The announcement highlights IREN’s successful pivot from bitcoin mining, now firmly establishing its role in providing crucial computing power for the rapidly expanding artificial intelligence sector. It’s a remarkable turnaround for a company that had faced recent investor skepticism.
IREN stock rallies on robust AI cloud contracts
IREN’s share price saw an immediate and positive reaction, climbing approximately 10% in premarket trading before reaching a 16% surge during Monday’s session. This sharp ascent signals renewed investor confidence in the company’s strategic direction.
The rally offers a welcome respite for IREN, whose shares had dropped 31% in the month leading up to this announcement. Investors had previously expressed concerns in early July 2026 over a new multi-year retention equity package for Co-Founders and Co-CEOs Daniel Roberts and Will Roberts.
That package granted the brothers an aggregate 6% stake, raising worries about potential share dilution. But the market now seems focused on IREN’s operational achievements, particularly its ability to secure major AI infrastructure agreements.
Expanding GPU infrastructure and developer partnerships
These new multi-year contracts, valued at $2.8 billion, aren’t just about the revenue. They strategically include customer prepayments covering roughly 45% of the associated GPU capital expenditure, which significantly reduces IREN’s net funding requirements for these deployments.
This financial model helps to mitigate the substantial capital demands involved in building and maintaining cutting-edge AI infrastructure. It also secures long-term revenue, with a weighted average contract term of approximately four years.
IREN has also broadened its customer base. The list now features industry heavyweights like Microsoft and NVIDIA, alongside innovative AI firms such as Perplexity, Figure AI, Together AI, Fluidstack, Fireworks AI, Fal AI, and Hume AI. An unnamed “new leading AI developer” also joined this expanding roster.
These clients are set to utilize IREN’s services across both bare metal and managed cloud offerings. Co-Founder and Co-CEO Daniel Roberts affirmed the company’s rapid growth, stating, “Our vertically integrated AI Cloud platform is scaling at pace.”
Roberts noted IREN has expanded its self-built AI Cloud capacity from about 3 megawatts (MW) to 480 MW being delivered this year. The company is targeting an even more ambitious 1.2 gigawatts (GW) by 2027, reflecting the persistent demand for high-performance graphic processing units.
Building on previous high-value agreements
The recent success follows a series of impactful deals. In November 2025, IREN secured a massive $9.7 billion AI cloud services contract with Microsoft, focusing on providing access to NVIDIA GB300 infrastructure.
That landmark agreement included a substantial 20% customer prepayment, setting a precedent for IREN’s current funding strategy. Then, in May 2026, the company signed a five-year AI infrastructure cloud services contract with NVIDIA itself, valued at $3.4 billion.
As part of that NVIDIA deal, IREN also issued NVIDIA a five-year right to purchase up to 30 million ordinary shares at $70 each. If fully exercised and conditions are met, this could represent a $2.1 billion investment from the chip giant, further intertwining their strategic interests.
IREN’s strategic pivot and financial strength
The new contracts have prompted IREN to revise its financial outlook for the year. It has now raised its 2026 Cloud Annualized Run-Rate Revenue (ARR) target from $3.7 billion to more than $4 billion.
Crucially, approximately 85% of this revised target is already under contract. This strong forward visibility in revenue generation offers significant stability in a dynamic market, demonstrating the efficacy of IREN’s strategic pivot from cryptocurrency mining to AI cloud infrastructure.
IREN began its shift to AI infrastructure in 2024, a move that mirrored other former crypto miners like CoreWeave and Hut 8. This transition allowed the company to capitalize on the booming demand for GPU computing power. While the bitcoin market has seen its ups and downs, IREN’s reorientation has paid dividends.
The company operates a vertically integrated model, overseeing everything from land acquisition to data center construction and operation. This approach helps IREN avoid external colocation fees, maintaining tighter control over costs and enhancing operational agility.
Access to low-cost, renewable power in regions such as British Columbia and West Texas, boasting around 3 GW of grid-connected capacity, further bolsters its competitive edge. This focus on sustainable energy aligns with growing environmental concerns surrounding energy-intensive AI operations.
As of June 30, 2026, IREN also reported a robust cash and cash equivalents position of approximately $7.6 billion. This substantial liquidity provides a strong foundation for continued expansion and investment in its ambitious AI infrastructure projects.
Navigating the complex neocloud market
IREN operates within the “neocloud” sector, a specialized segment of the cloud computing market tailored for AI workloads. These providers focus on GPU-as-a-Service (GPUaaS), offering infrastructure specifically optimized for large-scale AI model training and high-performance data environments.
This specialization sets neoclouds apart from traditional hyperscalers like Amazon Web Services or Microsoft Azure, which offer a broader suite of generalized services. The targeted approach often leads to more predictable, flat-rate pricing and can be more cost-effective for highly demanding AI tasks. But this market isn’t without its challenges.
Recent weeks have seen concerns about potential overinvestment in neoclouds, especially with Meta Platforms reportedly planning its own expansion into cloud services. Such moves by hyperscalers to develop in-house data centers could intensify competition for specialized providers.
Building AI-ready infrastructure also demands significant capital, requiring massive investments in specialized processors, high-bandwidth memory, and advanced cooling systems. Beyond hardware, the energy consumption required to train and run large AI models is astronomical.
This places immense pressure on existing power grids and infrastructure, creating ongoing challenges for providers to secure reliable, affordable, and sustainable energy sources. These global market dynamics underscore the unique pressures facing specialized AI infrastructure firms.
Ambitious expansion for AI compute capacity
Looking ahead, IREN has outlined ambitious plans for the continued expansion of its physical infrastructure. This includes accelerating construction of its Horizon 1-4 liquid-cooled data centers in Childress, Texas, specifically designed for Microsoft and offering 200 MW of critical IT load.
The company is also developing a massive 1,800-acre, 2-gigawatt AI data center in Sweetwater, Texas. Substation energization for the first phase, Sweetwater 1 (1,400 MW), is targeted for April 2026, with Sweetwater 2 (600 MW) following in late 2027.
IREN’s consistent success in securing these substantial new contracts, particularly with customer prepayments, could set a precedent for financing large-scale AI infrastructure projects. This model helps de-risk capital expenditures and ensures predictable revenue streams, contrasting with broader market shifts.
The ongoing, voracious demand for GPU infrastructure continues to outstrip available supply, creating a highly favorable market for companies like IREN. Its ability to broaden its customer base beyond hyperscalers to include diverse enterprises and AI developers further strengthens its market position.