Category: AI Industry Updates

Disclosure: TopAINest is reader-supported. This post contains affiliate links, and we may earn a small commission if you buy through them — at no extra cost to you. We only recommend products we’d genuinely tell a friend to buy.

In January 2025, OpenAI, SoftBank, and Oracle stood alongside President Trump at the White House to announce Project Stargate — a joint venture promising to invest up to $500 billion in AI infrastructure across the United States over four years. It was the kind of announcement that made even seasoned tech observers pause. Half a trillion dollars. For data centers.

Since then, the project has evolved, attracted additional partners, broken ground on its first campuses, and generated a fair amount of controversy. If you’ve seen the name in AI headlines and wondered what exactly is being built — and why it matters — this guide breaks it down in plain language.

Project Stargate is not just a single data center or a funding round. It is the largest private infrastructure investment in AI history, and it is betting that whoever controls the most powerful computing clusters will largely shape what artificial intelligence looks like in the next decade.

What Is Project Stargate?

Project Stargate is a joint venture created to build artificial intelligence infrastructure — primarily massive data centers packed with GPUs and specialised AI chips — across the United States. The name refers to the initiative as a whole, not a single facility. It is an umbrella project for what the founders describe as the physical foundation that future AI systems will require to operate at scale.

The core insight driving it is simple: training and running frontier AI models — the kind powering ChatGPT, advanced medical research tools, and autonomous systems — requires enormous amounts of computing power. Today’s data centers were not built with that demand in mind. Project Stargate aims to build the ones that are.

The first Stargate campus broke ground in Abilene, Texas, in early 2025. As of late 2026, construction is underway at multiple US sites, with Texas serving as the main hub. The project envisions up to 20 data center campuses in the US, with potential international expansion in later phases. Each campus is designed not as a single building but as a sprawling multi-building complex, some covering hundreds of acres.

Who Is Behind It and How Much Are They Spending?

The founding partners of Project Stargate are OpenAI, SoftBank, and Oracle. SoftBank committed to serving as the primary financial backer, with CEO Masayoshi Son taking the lead investor role. OpenAI provides the AI expertise and uses the infrastructure for its own model development. Oracle is handling a significant share of the data center construction and ongoing management.

Microsoft, which has been OpenAI’s largest investor since 2019 and provides cloud capacity through Azure, is not a founding partner of the Stargate JV — a detail that attracted notice when the project was announced. Microsoft has clarified that its own multi-billion-dollar AI infrastructure buildout continues independently, and the two relationships are complementary rather than competing. In practice, OpenAI continues to use both Microsoft Azure and the new Stargate facilities.

Other companies have joined as technology partners, meaning they supply components and expertise rather than capital. These include NVIDIA, Arm, and Oracle’s infrastructure subsidiary. NVIDIA’s role is particularly central: Stargate data centers are built around clusters of NVIDIA’s most advanced chips, including the Blackwell architecture GPU family announced in 2024 and its successors.

The $500 billion figure is the total investment target over four years. The first phase committed an initial $100 billion from SoftBank. Whether the full amount materialises depends on fundraising momentum, AI demand growth, and the pace at which the infrastructure generates returns. Even at half the stated target, the project would represent a historically unprecedented injection into US computing infrastructure.

What Will the Infrastructure Actually Look Like?

Stargate facilities are not conventional data centers. They are designed from the ground up for AI workloads, which differ from ordinary cloud computing in a critical way: AI training and inference are almost entirely GPU-bound. A single large-scale AI training run can consume tens of thousands of high-end GPUs running simultaneously for weeks or months. That kind of workload requires fundamentally different power density, cooling, and networking than a typical web-hosting facility.

The power demands are extraordinary. Some individual Stargate campuses are expected to draw one gigawatt or more at peak load — comparable to the power consumption of a mid-sized city. Securing grid access, building dedicated substations, and in some cases contracting directly with power producers has become one of the defining logistical challenges of the project. Some sites are exploring co-location with natural gas generation or dedicated renewable capacity to reduce dependence on existing grid infrastructure.

Cooling is the other major engineering challenge. Modern AI data centers typically use evaporative cooling to manage chip temperatures, which requires large volumes of water. Stargate has faced scrutiny on this front, particularly in West Texas where water scarcity is a real concern. Some campuses are piloting dry-cooling and closed-loop alternatives, though these approaches carry their own trade-offs at gigawatt scale.

On the compute side, the campuses are engineered around NVIDIA’s NVLink interconnect fabric, which allows thousands of GPUs to communicate at extremely high bandwidth. This is a prerequisite for training models at the parameter scales that frontier AI labs are targeting in the coming years — models far larger than what current infrastructure can handle.

Why Does This Matter for the Global AI Race?

The competition for AI leadership is, at its foundation, a competition for computing infrastructure. A country or company with more powerful training clusters can develop more capable models faster, iterate more quickly, and — depending on how these capabilities develop — gain significant advantages in science, defence, economic productivity, and technology exports.

China has been building its own AI infrastructure aggressively, though US export controls on advanced semiconductors have constrained its access to the most capable chips. Project Stargate is partly a strategic bet that maintaining a domestic lead in AI compute — built on the most advanced chips produced in the US and its allied chip-fabrication partners — gives the US a durable advantage even as algorithmic development becomes more globally distributed.

It also matters commercially. OpenAI’s ability to train the next generation of models, scale ChatGPT to its hundreds of millions of users, and develop new AI products depends directly on having access to this kind of infrastructure. Without it, compute becomes the ceiling on what is possible, and that ceiling arrives quickly in modern AI development. The same logic applies to every AI company that depends on cloud inference capacity: the underlying infrastructure is the foundation everything else is built on.

Concerns, Criticism and Open Questions

Project Stargate has attracted meaningful criticism from several directions. The most prominent early challenge came from Elon Musk, who publicly claimed — without presenting supporting evidence — that SoftBank lacked the committed capital to back the deal. SoftBank and OpenAI disputed this; construction proceeding on schedule in Abilene was the most direct rebuttal. Musk has his own AI venture (xAI) and a competing infrastructure buildout (the Colossus supercomputer cluster in Memphis), so the competitive context of his comments is relevant.

Substantive critics have focused on environmental impact. The energy demands of multi-gigawatt AI campuses are significant, and there are legitimate questions about whether these facilities are competing for grid capacity that would otherwise serve residential or industrial users. Proponents argue that AI-driven efficiency gains across industries could offset the energy cost many times over — plausible in principle, though unverified at this scale. The water consumption question is harder to deflect, particularly in drought-prone regions.

There are also execution risks. Building 20 major data center campuses in four years is a massive logistical undertaking. Permitting, grid interconnection, construction labour, and chip supply chains all introduce potential delays. And as a joint venture between three large companies with sometimes diverging commercial interests, the governance of Stargate as it matures will be worth watching closely — particularly if AI market conditions shift significantly from what the founders projected in 2025.

Further Reading

If you want to go deeper on the forces driving projects like Stargate, these books offer strong grounding in the economics, geopolitics, and risks of the AI build-out:

• The Coming Wave by Mustafa Suleyman — a co-founder of DeepMind examines why AI and synthetic biology are among the most powerful and difficult-to-contain technologies in human history. See on Amazon → (~$20)
• Power and Progress by Daron Acemoglu and Simon Johnson — two economists argue that AI’s benefits are not inevitable and depend heavily on who controls the infrastructure and the incentives shaping it. See on Amazon → (~$25)
• The Age of AI by Henry Kissinger, Eric Schmidt, and Daniel Huttenlocher — a wide-angle view of how AI is reshaping geopolitics, economics, and human cognition from three writers with very different vantage points. See on Amazon → (~$18)

Frequently Asked Questions

The Bottom Line

Project Stargate is the largest private infrastructure investment in AI history. It reflects a genuine belief — among some of the world’s most powerful technology organisations — that the physical layer of AI: the chips, the power, the cooling, the fiber, is going to be as strategically important as the models running on top of it.

Whether the full $500 billion gets deployed, whether the timelines hold, and whether the environmental trade-offs are handled responsibly are all legitimate open questions. What is not in question is the direction: AI is becoming an infrastructure industry, and Project Stargate is the boldest bet yet on what that infrastructure will need to look like.

If you are following the AI industry in 2026, understanding what is being built at the physical layer — and who is building it — is becoming as important as understanding the models themselves.

Disclosure: TopAINest is reader-supported. This post contains affiliate links, and we may earn a small commission if you buy through them — at no extra cost to you. We only recommend products we’d genuinely tell a friend to buy.

The race to build the world’s most powerful supercomputers has never moved faster. In 2026, the machines at the top of the list aren’t just research tools — they’re the engines driving AI breakthroughs, drug discovery, climate modelling, and national security. Here’s where things actually stand.

Why Supercomputers Matter More Than Ever in 2026

The connection between supercomputers and AI is now inseparable. Training frontier AI models — the kind that power ChatGPT, Gemini, and their successors — requires exaflop-scale compute. A single large model training run can consume more computing power than all scientific computing done globally just a decade ago. Countries and corporations are competing on supercomputer power as a direct proxy for AI capability, and the stakes are geopolitical.

The Current Top 5 Supercomputers (2026)

1. Frontier — Oak Ridge National Laboratory, USA

Performance: 1.2 exaflops (1.2 quintillion calculations per second)
Builder: HPE Cray / AMD
Primary use: Climate modelling, nuclear stockpile simulation, materials science, AI research

Frontier, at Oak Ridge National Laboratory in Tennessee, was the world’s first exascale computer when it launched in 2022 and remains a top-tier system in 2026. It uses AMD EPYC CPUs and AMD Instinct MI250X GPUs — the same GPU architecture that powers enterprise AI workloads. Frontier has contributed to breakthroughs in protein folding predictions, fusion energy research, and extreme weather modelling. It consumes about 21 megawatts of power — roughly equivalent to 21,000 Canadian homes.

2. Aurora — Argonne National Laboratory, USA

Performance: 2+ exaflops (targeting)
Builder: HPE Cray / Intel
Primary use: AI for science, particle physics, cancer research

Aurora at Argonne National Laboratory is the U.S. Department of Energy’s second exascale supercomputer, built around Intel’s Ponte Vecchio GPUs. It was designed explicitly for AI-driven scientific discovery — running large-scale machine learning models alongside traditional simulation workloads. Aurora has been used for projects including modelling the human brain at cellular scale and accelerating drug candidate screening. Its Intel GPU architecture is distinct from Aurora’s Nvidia/AMD competitors, making it a proving ground for Intel’s datacenter AI ambitions.

3. El Capitan — Lawrence Livermore National Laboratory, USA

Performance: ~2 exaflops
Builder: HPE Cray / AMD
Primary use: Nuclear weapons stockpile stewardship (classified), AI

El Capitan, delivered to Lawrence Livermore in late 2024, was built primarily for nuclear weapons simulation — ensuring the U.S. stockpile remains safe, secure, and effective without physical testing. It uses AMD EPYC Genoa processors and AMD Instinct MI300A accelerators, the first processor to combine CPU and GPU memory in a single chip. While much of its work is classified, El Capitan is confirmed to be among the fastest systems in the world and represents the bleeding edge of what U.S. national labs can deploy.

4. MareNostrum 5 — Barcelona Supercomputing Center, Spain

Performance: ~314 petaflops general-purpose + exascale GPU partition
Builder: HPE / Intel + Nvidia
Primary use: European research — climate, genomics, AI

MareNostrum 5, housed in a stunning converted chapel in Barcelona, is Europe’s flagship supercomputer. It features a mixed architecture — Intel CPUs for general-purpose workloads and Nvidia H100 GPUs for AI-specific tasks. The Barcelona Supercomputing Center uses it to run the EU’s flagship AI initiatives, climate modelling for the Copernicus programme, and genomics research across European universities. The iconic setting (server racks in a 19th-century church nave) makes it one of the most photographed data centres in the world.

5. Fugaku — RIKEN, Japan

Performance: ~442 petaflops
Builder: Fujitsu / ARM
Primary use: COVID-19 research legacy, climate, disaster simulation

Fugaku, at the RIKEN Centre for Computational Science in Kobe, held the #1 spot on the TOP500 list from 2020–2022. Built on Fujitsu’s A64FX ARM-based processor, it’s unique as the only top-tier supercomputer built on ARM architecture. Fugaku played a critical role in COVID-19 droplet simulation research that shaped mask guidance and social distancing policy in Japan. While newer exascale machines have surpassed its raw performance, Fugaku remains in the top ten globally and continues to run critical national research programs.

What’s Coming: The Next Generation (2026–2028)

The next wave of supercomputers will be measured in zettaflops — 1,000 exaflops. Several projects are already announced:

• JUPITER (Germany) — A new European system targeting 1 exaflop, emphasizing energy efficiency
• ISAMBARD-AI (UK) — The UK’s largest AI-focused supercomputer, Nvidia GH200 Grace Hopper Superchips
• Stargate (USA) — OpenAI and SoftBank’s $500 billion AI infrastructure investment, which includes building next-generation supercomputing clusters for frontier AI training
• China — Multiple classified exascale projects believed to be operational; rarely reported in TOP500 due to export control concerns

How to Learn More — Books and Resources

📘 High Performance Computing: Modern Systems and Practices — the standard textbook for understanding how supercomputers are architected, programmed, and run. Used in graduate CS programs worldwide.

📘 Chip War by Chris Miller — the definitive account of the global semiconductor competition that underlies the supercomputer race. A must-read for understanding why compute power is now a geopolitical weapon.

📦 Nvidia Jetson Nano Developer Kit — run your own GPU-accelerated AI workloads on a $99 board. The same CUDA architecture that powers Frontier and Aurora, scaled to your desk.

FAQ

What is the fastest supercomputer in the world in 2026?

As of 2026, Frontier at Oak Ridge National Laboratory remains among the verified fastest at 1.2 exaflops. El Capitan and Aurora are both operational and likely comparable or faster, but their exact benchmarks are less publicly documented. China may have unreported systems that rival or exceed these figures.

What is an exaflop?

One exaflop equals one quintillion (10¹⁸) floating-point operations per second. To put that in perspective: if every person on Earth did one calculation per second, it would take about 4 million years to match what an exascale computer does in a single second.

Why does Canada not have a top-10 supercomputer?

Canada has strong HPC infrastructure — the Digital Research Alliance of Canada operates Narval, Béluga, Cedar, and Graham clusters — but none approach exascale performance. Funding for a Canadian exascale system has been discussed but not yet committed. The country’s AI talent and research output exceeds what its domestic compute infrastructure would suggest, largely because Canadian researchers use U.S. and European facilities.

The Bottom Line

Supercomputers in 2026 are the infrastructure of the AI revolution. The countries and institutions that control exascale compute have a meaningful advantage in AI research, drug discovery, materials science, and national security simulation. The race to build faster, more energy-efficient systems is accelerating — and the machines coming in 2027–2028 will make today’s top systems look modest.

For the rest of us, understanding what these systems can do — and what they’re being used for — is increasingly important context for understanding the AI breakthroughs making headlines every week.