Nvidia and IREN join forces to install 5 GW of infrastructure dedicated to artificial intelligence

Faced with the explosion of computing needs related to artificial intelligence, two giants of the technology sector have joined forces to transform the infrastructure landscape. Nvidia, global leader in graphics processors, is partnering with IREN, a major player specialized in the operation and development of data centers, to deploy up to 5 gigawatts (GW) of infrastructure dedicated to artificial intelligence. This major collaboration aims to accelerate the construction of a new generation of AI factories capable of supporting the massive loads generated by advanced AI applications, particularly in a context where demand for computing power and energy is unprecedented.

The partnership combines Nvidia’s technical expertise in GPU and AI architecture with IREN’s mastery of integrated management of energy, land, and operational infrastructures. Thanks to this strategic union, the goal is not only to increase installed capacity through a global network of green data centers but also to rethink how technology and energy innovation come together to meet the demands of a rapidly evolving sector. The Texas site of Sweetwater is already emerging as the nerve center of this program, embodying the vision of industrial, reliable, and sustainable AI factories.

The exceptional context of scaling AI infrastructures up to 5 GW

By 2026, the dynamic of technological innovation is radically transforming the infrastructure needs of artificial intelligence. Advances in generative AI, with its applications in natural language processing, computer vision, and multimodal conversational agents, have sharply increased the demand for intensive computing. This explosion involves not only more GPUs but also greater energy, sophisticated data centers, and adapted networks capable of supporting these enormous workloads.

In this context, the partnership between Nvidia and IREN represents a strategic response to these challenges. The envisaged capacity of 5 gigawatts to be deployed will equip data centers at a global scale, enabling rapid and sustainable scaling of AI infrastructures. This threshold is symbolic: it demonstrates the two companies’ intention to build industrial AI factories capable of hosting massive GPU clusters necessary for training and inference of advanced machine learning models.

Beyond raw power, it also involves ensuring coherent integration between energy supply, land management, and the technical deployment of AI architectures. This innovative industrial approach breaks with one-off constructions by favoring a vertically integrated model, where mastery of the entire value chain is at the heart of success.

The Sweetwater campus, located in Texas, thus becomes an emblematic example. With a planned energy capacity of 2 GW, this site will benefit from cutting-edge infrastructures, combining privileged access to renewable energy and a favorable technology ecosystem. It is thus poised to become a key hub for global AI innovation, meeting environmental standards while offering unprecedented computing performance.

Finally, this partnership illustrates a paradigm shift where the race for computing power is no longer limited to processor performance but now incorporates the energy dimension and sustainable resource management. This new model establishes itself as essential for an AI industry that strives to combine efficiency, scalability, and respect for climate objectives.

Energy and technological challenges related to large-scale AI infrastructures

The power required to operate rapidly expanding artificial intelligence models is not limited to the quantity of GPUs deployed. It also requires a reliable, sustainable, and large-scale energy supply. The Nvidia-IREN partnership integrates this reality by prioritizing efficiency and energy sobriety in the deployment of its infrastructures.

Traditional data centers, often oversized to handle variable loads, become obsolete in the face of the specific demands of AI factories. These require stable power supplies, advanced thermal management, and precise resource control to avoid losses and optimize electricity consumption.

IREN has unique expertise in this field, combining not only mastery of physical infrastructures but also integration of renewable energy to power its data centers. This energy choice is part of the broader goal of a responsible industry capable of supporting the digital revolution without sacrificing the planet.

Concrete examples of energy innovation in Sweetwater

In Sweetwater, a comprehensive ecosystem is being implemented to guarantee the energy reliability of the future AI factory:

• Integration of wind and solar farms directly connected to data centers to maximize the use of renewables.
• Intelligent energy management systems combining battery storage and dynamic resource allocation according to load.
• Greening of infrastructures through low-carbon footprint materials and innovative cooling techniques.
• Optimized consumption management, notably by leveraging artificial intelligence algorithms to anticipate and reduce consumption peaks.

These measures illustrate to what extent the technological alliance between Nvidia and IREN goes far beyond simple calculations: it positions energy as an inseparable component of AI performance. This strategy is now essential to attract native AI companies as well as cloud giants seeking to secure their resources in a global energy transition context.

Simultaneous adaptation of infrastructures and energy management will be a key lever to allow these AI factories to operate at full capacity without compromising environmental standards. This challenge highlights the growing complexity of the industry and the necessary coordination of multiple expertise areas.

An innovative DSX architecture to industrialize AI on a very large scale

The DSX architecture, developed by Nvidia, forms the technological core of the partnership with IREN. This infrastructure was expressly designed to respond to the scaling up of industrial AI deployments, aligned with the commitment to technological innovation and systemic optimization.

The DSX is a modular platform that combines high-performance GPUs, fast interconnections between servers, and hardware configurations adapted to different phases of model training and inference. The challenge is to enable smooth scalability and efficient resource management while minimizing operating costs.

Thanks to this solution, IREN’s data centers will progressively host massive AI clusters capable of delivering tens of petaflops or more, necessary for the most complex computational tasks. The DSX is also optimized to reduce latency, which is crucial for real-time applications and innovative cloud services.

Adaptability and modularity serving future needs

The interest of a technology like the DSX also lies in its ability to evolve with AI advances. New generations of GPUs or storage technologies can be integrated without major disruption, ensuring the longevity of investments. It also offers flexibility to meet the diversity of AI applications, whether dedicated to language processing, visual recognition, or predictive modeling.

This architecture allows Nvidia and IREN to secure a competitive advantage in the global race for computing power. By offering a comprehensive infrastructure, designed from the outset to accommodate exponential growth in demand, they position their data centers as “AI factories” capable of supporting rapid market evolutions.

The economic and strategic implications of the Nvidia-IREN partnership

Beyond the technical aspect, this partnership also marks a turning point in the economic and industrial strategy of both players. Nvidia is no longer just a GPU supplier; the company becomes a full-fledged integrator of AI infrastructures on a global scale. From its side, IREN leverages this alliance to strengthen its global presence and its positioning as an expert in comprehensive solutions, ranging from energy management to computing capacity provision.

The deal includes a significant financial component: Nvidia has a purchase right of up to 30 million ordinary IREN shares over five years, at a price set at 70 dollars per share. This option could raise Nvidia’s total investment to 2.1 billion dollars, reflecting strong confidence in the viability and potential of this industrial partnership.

In a highly competitive global market, mastering simultaneously the hardware, energy, and operational aspects of AI infrastructure is a real differentiating advantage. It enables them to reduce commissioning times, contain costs, and ensure a rapidly scalable and reliable service.

Effects on the global AI market

This collaboration will significantly influence the dynamics of hyperscalers and cloud providers. It positions IREN as an essential partner, capable of offering robust solutions to host and operate the most advanced AI technologies, addressing the needs of startups, large companies, and digital service providers.

This integrated infrastructure model could be replicated on other continents, thus contributing to the emergence of a global network of AI factories. By securing both hardware and energy, Nvidia and IREN define a new standard in deployment, with a lasting impact on global competitiveness.

Sweetwater: future American capital of AI infrastructures

The choice of Texas to establish the Sweetwater campus is not accidental. This region attracts significant investments thanks to favorable conditions: extensive land, access to abundant and renewable energy, as well as good network connectivity.

The Sweetwater campus, with a planned capacity of 2 GW, will become a real hub for the accelerated deployment of Nvidia’s DSX architecture. This site embodies IREN’s vision of innovation, which bets on vertical integration combining clean energy, data center management, and efficient exploitation of GPU clusters.

This initiative also illustrates the strategic importance Texas is gaining on the global AI map. Its geographical position and regulatory stability create an attractive environment for large technology companies seeking to secure their energy supply while maximizing computing performance.

Impact on AI players and customer companies

The AI factories developed under the partnership will enable artificial intelligence players to access massive computing capacity while benefiting from an optimized and environmentally friendly infrastructure. This increased availability will accelerate the development of AI applications in various sectors such as healthcare, finance, mobility, and scientific research.

Startups specializing in AI will also find a robust platform to test and train their models at scale without having to invest alone in costly and complex infrastructures to manage. Moreover, large companies will be able to outsource or extend their computing capacities by relying on cloud services aligned with the latest Nvidia and IREN technologies.

Furthermore, this scaling of infrastructures will reduce latency times and improve performance in real-time applications, an increasing requirement for many uses. This could accelerate the adoption of innovative solutions such as augmented reality, intelligent virtual assistants, and automated driving systems.

Finally, the standardization of this DSX architecture within IREN data centers will contribute to harmonizing resource management and simplifying operations. This will benefit the entire AI ecosystem by enabling faster and more reliable deployment of new solutions.

List of key benefits for end users:

• Scalable computing capacity adapted to their specific needs.
• Access to cutting-edge infrastructure with guaranteed performance.
• Energy optimization thanks to the integration of renewable energies.
• Reduced deployment times for large-scale AI projects.
• Support for startups and easier access to resources for innovation.
• Reduced environmental impact compared to traditional data center models.

Perspectives for global expansion and influence on the future of AI infrastructures

Building on this unprecedented American-scale project, Nvidia and IREN aim to roll out their model across several regions worldwide. Their infrastructures are being deployed in strategic areas in Europe, Asia-Pacific, and North America, through centers already operational or in planning stages.

The potential is considerable, notably because the need for innovation in energy design and logistics management continues to grow. AI models are becoming more complex, with increased demands for multimodal and real-time solutions. Without infrastructures up to the task, these advances could not be effectively exploited at large scale.

In this context, Nvidia and IREN are also developing local partnerships with public and private stakeholders to adapt their infrastructures to regional specificities, thus guaranteeing agility and resilience.

Region	Planned AI Capacity (GW)	Main Advantages	Progress Status
North America	3.0	Access to renewable energy, extensive land, dense network	Advanced implementation phase
Europe	1.2	Strict environmental standards, green data centers	Initial deployment
Asia-Pacific	0.8	Fast-growing markets, diversified energy access	Planning phase

With global deployment across several continents, this integrated infrastructure strategy aims to position Nvidia and IREN as key players defining tomorrow’s standards for industrial AI. By securing the complete chain between materials, energy, and operation, they offer a pragmatic response both technical and economic to upcoming challenges.

Towards a new era of artificial intelligence industrialization

The alliance between Nvidia and IREN fits within a logic clearly marked by the massive industrialization of artificial intelligence. This stage follows the period when competition was mainly focused on the design of models and algorithms.

Today, the real challenge lies in deploying large-scale technical infrastructures capable of supporting the explosion of uses while ensuring performance, reliability, and scalability. This industrialization translates into the construction of AI factories that resemble industrial complexes more than simple data centers.

In this context, the joint mastery by Nvidia and IREN of hardware, software, energy, and operational resources provides them with an integrated vision and reinforced capacity for action. The AI factories thus created become a fundamental foundation for the global digital economy, comparable to historical infrastructures such as electrical grids or telecommunications.

This transformation opens the way to even more disruptive innovations, where computing power becomes a widely accessible resource, enabling accelerated research, business applications, and democratization of AI uses.