Perspective ·
Michael Dell Surpasses Mark Zuckerberg: What the AI Infrastructure Boom Reveals
Dell’s latest results show how quickly AI infrastructure is scaling. The next question is how people will turn that compute into models they can understand, shape and run on their own terms.
On October 1, Forbes estimated Michael Dell’s net worth at $268 billion, up $28 billion over the month as Dell Technologies shares rose 18% in September. In that dated snapshot, Forbes placed Dell ahead of Mark Zuckerberg. Rankings and wealth estimates move with markets; the infrastructure investment behind the headline is visible in company results.
What the numbers say
Dell’s fiscal 2027 second-quarter results, released September 1, show the scale of demand for AI systems:
- $47 billion in quarterly revenue, up 58% year over year.
- $16.4 billion in AI-optimized server revenue, up 100%.
- $60.9 billion in AI server orders booked during the quarter.
- $95 billion in AI server backlog at quarter-end.
The company also raised its full-year fiscal 2027 guidance to $74 billion in AI-optimized server revenue, up 200%, and $192 billion in total revenue. These are company forecasts, not completed sales.
The distinctions matter: revenue is what Dell recognized during the quarter; orders are bookings; backlog is work awaiting fulfillment; and annual guidance is a projection. They show different parts of a fast-growing market, and should not be treated as interchangeable.
From infrastructure to AI ownership
The AI conversation has often focused on model capabilities, benchmarks and access to APIs. The growth in infrastructure brings another set of practical questions into view: where data is processed, what training will cost, how a model can be adapted, and where it can run.
Cloud services, hosted models and privately operated systems will continue to coexist. As organizations invest in AI compute, they also need tools that help them use it efficiently and choose how much control they want over their models and data.
The software layer NEURAX is building
NEURAX brings model design, hardware-aware analysis, dataset profiling, training and export into one studio. For supported workflows, the Studio Agent can turn a natural-language request into an architecture on the canvas, analyze its requirements against the selected hardware and explain the results before a run begins.
Datasets are profiled where they are stored. When an AI provider is used, the assistant receives a summary of the data, such as shapes and counts, rather than the file path or raw contents. You review the plan and approve training; runs produce ordinary project files, checkpoints and model weights that you can use in your own environment.
That is the role we want NEURAX to play: software that helps people turn computing resources, data and ideas into models they can inspect, train and carry with them. The Linux desktop release is available now; macOS and Windows support are in progress.
Our primary market is universities and companies around the world, where teams need to develop AI around their own data and computing resources. Individuals can use NEURAX too: students, researchers, educators and independent builders are welcome to work in the same studio.
NEURAX — Build Your AI. Run It Your Way. Explore NEURAX.
Sources: Forbes, October 2026 wealth estimates; Dell Technologies, fiscal 2027 second-quarter results; Dell Technologies, Form 10-Q.