Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

Lloyd Wamsley notes that frontier labs currently buy neocloud capacity out of sheer necessity during the compute rush. However, long-term economics dictate that major tech players will build and own their own facilities, leaving neoclouds vulnerable once private infrastructure goes online.

Related Insights

Firms like OpenAI and Meta claim a compute shortage while also exploring selling compute capacity. This isn't a contradiction but a strategic evolution. They are buying all available supply to secure their own needs and then arbitraging the excess, effectively becoming smaller-scale cloud providers for AI.

Investor Finn Barnes predicts that to achieve maximum efficiency for training frontier AI models, hyperscalers will shift away from multi-tenant infrastructure. Instead, they will build entire gigawatt-scale campuses dedicated to a single, massive customer, creating bespoke infrastructure to support the largest players.

At scale, renting compute from AWS, Google, or Microsoft is a strategic mistake for AI leaders like OpenAI and Anthropic. It creates a critical dependency, forcing them to enter the capital-intensive data center business to control their supply chain and destiny.

The primary bear case for specialized neoclouds like CoreWeave isn't just competition from AWS or Google. A more fundamental risk is a breakthrough in GPU efficiency that commoditizes deployment, diminishing the value of the neoclouds' core competency in complex, optimized racking and setup.

Instead of bearing the full cost and risk of building new AI data centers, large cloud providers like Microsoft use CoreWeave for 'overflow' compute. This allows them to meet surges in customer demand without committing capital to assets that depreciate quickly and may become competitors' infrastructure in the long run.

Investing tens of billions into proprietary compute clusters carries less downside risk when an organization owns the physical hardware. If internal consumer models underperform, valuable data center capacity can be redeployed internally or leased to compute-hungry third parties. Owning depreciable infrastructure also provides structural cost advantages over renting cloud compute.

Despite macroeconomic headwinds and debt costs, hyperscalers continue funding a forecast $31 trillion data center build-out because they see a fleeting window to control foundational infrastructure. Once global compute capacity is fully built, digital infrastructure will transition toward commoditization. Big tech operators are willing to absorb margin compression now to secure control of the physical pipes before market dynamics lock down.

The NeoCloud landscape is expanding beyond pure-play providers. Inference-focused software platforms and custom chip startups are now building their own data centers and buying GPUs. This vertical integration is driven by the need to control supply, meet demand, and improve margins, fundamentally changing their business models.

While costs for essentials like copper and electricity are rising, cash-rich hyperscalers (Google, Meta) will continue building. The real pressure will be on smaller, capital-dependent players like CoreWeave, who may struggle to secure financing as investors scrutinize returns, leading to canceled projects on the margin.

Leading AI firms like Anthropic are moving beyond flexible cloud consumption to securing massive, multi-year capacity contracts for private data centers. This shift to "capacity pre-emption" signals that guaranteed access to scalable infrastructure is now as critical an asset as the AI models themselves.