TheSequence

TheSequence

Share this post

TheSequence
TheSequence
The Sequence #707: Rise of the Neo-Clouds: Can Startups Beat the Cloud Giants in AI Compute?

The Sequence #707: Rise of the Neo-Clouds: Can Startups Beat the Cloud Giants in AI Compute?

The present and future of one of the most interesting categories in AI.

Aug 21, 2025
∙ Paid
13

Share this post

TheSequence
TheSequence
The Sequence #707: Rise of the Neo-Clouds: Can Startups Beat the Cloud Giants in AI Compute?
Share
Created Using GPT-5

This is a looooooooooong post but hopefully insightful. Lately I’ve been doing some research in the category of GPU-clouds known as neo-clouds which have become quite hot in recent years. Despite the amazing raise, there is a serious argument that a lot of these platforms are going to face significant pressure and even get comoditized. How to reconcile the two? I started putting together my ideas and ended up with a very long analysis.

The term neo-cloud refers to cloud platforms focused almost entirely on offering GPU-based computing as a service. In contrast to the general-purpose “hyperscale” cloud providers (like Amazon Web Services or Google Cloud) that offer hundreds of services, neo-clouds specialize in one thing: providing raw, high-performance compute for AI and other intensive workloads.

In this essay, we will explore the rise of these GPU-centric neo-cloud platforms and analyze their current state in the industry. We will examine the drivers behind their growth, including the insatiable demand for AI compute and the gaps left by traditional cloud giants. We’ll also discuss how the landscape might evolve toward a commoditized future and what that means for both upstart providers and established hyperscalers. A comparison with the hyperscalers is essential, especially as companies like Amazon and Google invest in building their own custom chips and GPUs to secure their dominance. We will then delve into the strategic paths neo-clouds might take to differentiate themselves—either by moving up the stack (offering integrated AI tools and services on top of hardware) or down the stack (optimizing the infrastructure for maximum performance and efficiency). Throughout, we will highlight examples of leading platforms in this space, and include a brief examination of geopolitical factors shaping the race for AI infrastructure. The tone will remain technical and analytical, yet engaging, to provide a comprehensive understanding of this fast-evolving domain.

The Emergence of GPU-Focused “Neo-Clouds”

This post is for paid subscribers

Already a paid subscriber? Sign in
© 2025 Jesus Rodriguez
Privacy ∙ Terms ∙ Collection notice
Start writingGet the app
Substack is the home for great culture

Share