Nvidia Invests $1 Billion in AI Startup Poolside Alongside $6 Billion Software Licensing Deal

Nvidia has struck a massive dual-pronged agreement with AI coding startup Poolside, injecting a combined $7 billion into the company through equity and licensing. The chipmaker is investing $1 billion into Poolside at a $12 billion pre-money valuation, while simultaneously paying $6 billion for a non-exclusive license to the startup's "Model Factory" software. The massive capital infusion fundamentally reshapes the trajectory of one of the industry's most closely watched artificial intelligence developers.[1][2]

The arrangement represents a novel structure in the artificial intelligence sector, functioning as a strategic partnership and partial talent acquisition rather than a full corporate buyout. As part of the agreement, Nvidia is extending job offers to approximately 109 Poolside employees—primarily the core engineers and researchers who built the startup's Laguna family of coding models.[1][3]

Despite the significant transfer of engineering talent, Poolside will continue to operate as an independent entity. Co-founders Jason Warner, the former chief technology officer at GitHub, and Eiso Kant will remain at the helm of the startup to pursue independent research initiatives. The $6 billion licensing fee, an unusually large sum for a non-exclusive software agreement, is expected to be distributed to Poolside's existing investors by the end of 2027, effectively providing an exit-level payout while keeping the company intact.[1][2]

For Poolside, the Nvidia deal serves as a critical lifeline following a turbulent fundraising period. Late last year, the startup attempted to raise $2 billion at a $12 billion valuation specifically to secure a massive cluster of 40,000 Nvidia GB300 chips. When the financing failed to materialize in time, Poolside lost access to the hardware, underscoring how access to raw compute power has become the ultimate bottleneck for mid-tier AI companies attempting to train frontier models.[1][2]

Nvidia's acquisition of Poolside's model-building machinery and engineering talent directly feeds into its "Nemotron" project, a family of open-weight AI models. The engineers transitioning to Nvidia will be tasked with integrating Poolside's software infrastructure into the Nemotron ecosystem, accelerating the chipmaker's ability to release highly capable models that developers can download and modify freely.[2][3]

By bolstering its internal software capabilities, the hardware giant aims to build a robust American open-weight ecosystem. This initiative is designed to compete directly with highly capable, low-cost Chinese models like DeepSeek and Kimi, which have rapidly gained market share among developers. It also positions Nvidia as a formidable alternative to proprietary US offerings from major labs like OpenAI and Anthropic.[3]

Market analysts view the move as a strategic wager by Nvidia to ensure the continued proliferation of open-source artificial intelligence. Widely available, highly capable open models drive broader enterprise AI adoption and lower the barrier to entry for software developers, which in turn fuels the underlying demand for Nvidia's core data center infrastructure.

Furthermore, by licensing Poolside's technology rather than acquiring the company outright, Nvidia navigates the tightening antitrust scrutiny surrounding Big Tech acquisitions. The deal mirrors similar arrangements Nvidia has struck with other AI startups, allowing the hardware giant to secure critical intellectual property and specialized talent without triggering the regulatory hurdles associated with traditional mergers.[1]