Neocloud
Specialized GPU-focused cloud providers that rent AI compute capacity, distinct from general-purpose hyperscalers.
As defined in Dan Sundheim - The Art of Public and Private Market Investing - [Invest Like the Best, EP.460]
Other definitions
A GPU cloud provider focused on renting AI compute, often financed by contracts signed with customers.
AI Market Jitters
Where it comes up · 9
- Inside the Trillion-Dollar AI Buildout7 claims
- Martin Casado on Frontier Model Economics, AI's Slowdown, and the Cursor Deal - [Invest Like the Best, EP.498]3 claims
- Ramtin Naimi - The State of Venture - [Invest Like the Best, EP.499]3 claims
- AI Market Jitters2 claims
- Ben Thompson on Big Tech, China, and the AI Boom Running Out of Money - [Invest Like the Best, EP.487]2 claims
- Eric Vishria - A Decade of Lessons Investing in Software & Hardware - [Invest Like the Best, EP.486]2 claims
- Dwarkesh Patel Podcast - Dylan Patel – Anthropic & OpenAI will have most of the world’s compute by 20281 claim
- Dan Sundheim - The Art of Public and Private Market Investing - [Invest Like the Best, EP.460]1 claim
and 1 more
Most important claims · 10 of 22Search all claims for “Neocloud” →
- Many investors early on refused to invest in AI models because they would never accrue value or be defensible, yet in a large growing market the model layer accrues substantial value regardless, just as neoclouds did despite poor revenue quality.Martin Casado on Frontier Model Economics, AI's Slowdown, and the Cursor Deal - [Invest Like the Best, EP.498]
- Contracted installed compute trades at a massive discount to the current spot market, because neoclouds signed long-term offtake contracts to finance GPUs; as those contracts roll off and reprice higher, operating cash flow should accelerate even if spot prices later decline.AI Market Jitters
- Nvidia's margins appear intact, but its circular financing, including a reported 25 percent backstop and equity stakes in neoclouds that guarantee compute purchases through 2030, lowers its customers' cost of capital and effectively acts as a hidden price cut.Ben Thompson on Big Tech, China, and the AI Boom Running Out of Money - [Invest Like the Best, EP.487]
- Neocloud GPU rentals earn high margins on short contracts initially, but short-term rentals become a terrible model as next-generation chips arrive at lower cost per performance, so long contracts with creditworthy counterparties such as Microsoft are the viable path.Inside the Trillion-Dollar AI Buildout
- Nvidia does not sell tokens or neocloud compute directly because it is strategically more valuable to cultivate a diverse community of neoclouds and inference providers that compete to create demand for Nvidia chips, with CoreWeave as an example.Neil Movva - Making AI 10x Cheaper - [Invest Like the Best, EP.488]
- Most compute is contracted before it is built, so it must be financed, and companies with balance sheets like Meta and SpaceX can build speculative capacity and sell to labs at high margins, creating a power structure different from neoclouds that need a customer and capital before building.Dwarkesh Patel Podcast - Dylan Patel – Anthropic & OpenAI will have most of the world’s compute by 2028
- Hyperscalers like Google and Amazon have a lower cost of capital than neoclouds, so in the capital fight they will outspend them, and they are also building chips they can sell externally, threatening Nvidia's margins as TPUs and Trainium become commodity alternatives.Ben Thompson on Big Tech, China, and the AI Boom Running Out of Money - [Invest Like the Best, EP.487]
- Fireworks runs the same open-source models on the same NVIDIA hardware as AWS, Azure, and GCP with about 5x better speed and a multiple-x throughput advantage, even while paying the cloud providers' margins, showing that running large models efficiently is very hard specialized expertise rather than commodity pass-through resale.Eric Vishria - A Decade of Lessons Investing in Software & Hardware - [Invest Like the Best, EP.486]
- Almost no one in 2024 or 2025 expected the price of old GPUs to rise sharply; the consensus assumed slow or steep price declines, so the current rise surprised the market.AI Market Jitters
- Nvidia is using its balance sheet through demand guarantees and cluster backstops, such as backing a CoreWeave cluster, because neoclouds will not build for short-term rental demand, which is anti-competitive in effect.Inside the Trillion-Dollar AI Buildout