Oracle, NVIDIA, AMD, SoftBank, and OpenAI setting the pace.
Whenever a small group of dominant companies gains control of an emerging industry, competitive risks are inevitable. In artificial intelligence, the race for supremacy in AI and AI infrastructure has fueled a surge of acquisitions, strategic partnerships, and massive investments for more than a decade. Today, that trend has accelerated dramatically: a handful of Fortune 500 firms and global investors have pledged nearly one trillion dollars to a few flagship projects and players — especially the market leading ChatGPT company OpenAI and the industry-backed infrastructure project Stargate — that are shaping the future of AI development.
Offering artificial intelligence to the world takes a lot of juice. Today, Oracle has emerged as OpenAI’s primary infrastructure partner in a $300 billion cloud computing agreement that will supply 4.5 gigawatts of data center capacity over five years starting in 2027. This deal is part of the Stargate initiative, a $500 billion AI infrastructure project launched in January 2025 with backing from Oracle, OpenAI, SoftBank, and others.
Oracle’s cloud infrastructure (OCI) will support OpenAI’s next-generation models, including GPT-5 and beyond. The partnership includes five new data centers in Texas, New Mexico, Ohio, and the Midwest, with Oracle overseeing construction and financing.
NVIDIA has committed up to $100 billion to OpenAI in a deal that will deploy 10 gigawatts of GPU-powered compute capacity using its Vera Rubin platform. The first gigawatt is expected to come online in late 2026.
This partnership gives NVIDIA a preferred role in OpenAI’s infrastructure roadmap and includes deep technical integration. Analysts have noted the circular nature of the deal: OpenAI receives capital from NVIDIA, which it then uses to purchase NVIDIA chips (Source: CNBC).
Oracle and NVIDIA are playing very different roles in OpenAI’s infrastructure strategy, which explains the disparity between investment size and gigawatt capacity. Oracle’s $300 billion deal focuses on building and operating five massive cloud data centers that will provide 4.5 gigawatts of general-purpose capacity over five years. This includes power for servers, networking, cooling, and mixed workloads—not just AI compute. In contrast, NVIDIA’s $100 billion commitment is tightly concentrated on GPU-powered compute, delivering 10 gigawatts of raw processing power through its Vera Rubin platform. GPUs consume far more electricity per unit than CPUs, so even with a smaller dollar figure, NVIDIA’s footprint looks larger in gigawatts. Essentially, Oracle is the backbone—real estate, construction, and cloud services—while NVIDIA is the engine, supplying the specialized hardware that drives AI performance.
AMD Joins the Race
In a surprise move, OpenAI also struck a multi-generational partnership with AMD, committing to deploy 6 gigawatts of AMD Instinct GPUs starting in 2026. The deal includes warrants for up to 160 million AMD shares, potentially giving OpenAI a 10% stake in the chipmaker (Source: MSN). AMD expects the deal to generate tens of billions in revenue, positioning it as a serious competitor to NVIDIA in the AI hardware space (Source: Business Insider).
Microsoft’s Evolving Role
Microsoft remains a major investor in OpenAI and retains exclusive rights to OpenAI’s API on Azure. However, its role in infrastructure has shifted. Under a revised agreement, Microsoft now has a right of first refusal rather than exclusivity on new compute capacity. This change enabled OpenAI to pursue deals with Oracle and others for Stargate. Microsoft continues to support OpenAI through Azure credits and product integration, including Copilot and enterprise deployments (Source: Microsoft Blog).
SoftBank’s Infrastructure Push
SoftBank is developing two Stargate data centers in Lordstown, Ohio and Milam County, Texas, with its affiliate SB Energy providing rapid-build infrastructure. These sites will add 1.5 gigawatts of capacity and are expected to be operational by 2026. SoftBank’s involvement underscores its commitment to AI infrastructure and complements its earlier $40 billion investment in OpenAI (Source: Investing.com).
A Few Words About Nuclear Investments
To meet the soaring energy demands of AI, other tech giants are turning to nuclear power. Microsoft partnered with Constellation Energy to revive the Three Mile Island plant, while Google is working with Kairos Power to secure 500 MW from small modular reactors by 2030. Amazon has inked deals with X-Energy, Energy Northwest, and Dominion Energy, investing $500 million to develop SMRs in Washington and Virginia. Meta signed a power purchase agreement with Constellation for the Clinton nuclear plant in Illinois, and OpenAI is exploring fusion energy through Helion, backed by CEO Sam Altman. These partnerships aim to secure stable, carbon-free baseload power to fuel AI’s exponential growth.
Market Implications
The Stargate initiative represents a seismic shift in AI infrastructure, with combined investments approaching $1 trillion. These deals are rapidly concentrating control over essential AI inputs—cloud capacity and advanced chips—among a handful of dominant firms. The key participants and their roles include:
1) OpenAI – The lead AI developer driving the need for hyperscale compute.
2) Oracle – Primary cloud infrastructure partner, committed to building and operating data centers and providing up to 4.5 GW of capacity through Oracle Cloud Infrastructure (OCI).
3) SoftBank – Major investor and energy partner, developing sites and providing capital and renewable power solutions.
4) NVIDIA – Strategic compute and networking partner, pledging up to $100 billion and supplying GPUs for 10 GW of AI compute via its Vera Rubin platform.
5) Microsoft – While not a construction partner, remains a key technology ally and OpenAI’s primary cloud provider (Azure), deeply integrated into the ecosystem.
6) Additional investors – Reports mention the UAE’s MGX fund and other institutional backers in earlier phases.
All of this raises significant antitrust concerns, including:
Market Concentration: Oracle, NVIDIA, and AMD’s consolidation of AI infrastructure threatens competition in both cloud services and chip supply, potentially resulting in a hardware duopoly and limiting opportunities for new entrants.
Circular Financing: The structure of these deals—where partners invest in OpenAI and OpenAI then uses those funds to purchase their products—heightens scrutiny under related-party transaction rules and may artificially inflate market values.
Barriers to Entry: The unprecedented scale of investment creates high barriers for new competitors, reinforcing the dominance of current hyperscalers and making it difficult for smaller players to compete on price or performance.
Exclusive Partnerships and Lock-In: Microsoft’s access to OpenAI models via Azure and Oracle’s massive cloud deal could further entrench market leaders, raising regulatory questions about unfair competition and potential lock-in effects for customers.
Consumer Impact: Concentration of compute power could lead to higher AI service costs, reduced consumer access, and opaque pricing, while allowing dominant firms to leverage their position across multiple markets. The FTC has flagged these risks as potentially harmful to competition and consumers.
Systemic Risk: The interconnected nature of these investments may amplify economic risks if the anticipated AI productivity gains do not materialize, drawing comparisons to the dot-com bubble’s collapse.
Echoes of a 19th Century Power Struggle
Just as the late 19th century “War of the Currents” between Westinghouse and Edison (1880s–1890s) shaped the future of electricity in America, today’s AI power race is defining the next era of innovation. Back then, the battle was over how to electrify a nation (AC or DC?); now, it’s about how to energize intelligence itself. With billions pouring into data centers, nuclear partnerships, and grid-scale energy deals, the infrastructure arms race for AI mirrors that historic struggle.
