You Asked for Our Recipe. Here It Is.

A weekly satellite-verified data center forecast is now live on the Vulcan UI — so you can stop guessing at market consensus and start building your outlook on ground truth.

From day one, Vulcan was built from an analyst's mindset, not a technologist's. Our job was to hand you the ingredients - the monitored projects, the milestones, the satellite ground truth - so you could develop a sharper forecast, whether you're tracking power or the data centers that consume it.

The request we heard most often was simple: "Apply your own analyst mindset. Use your ingredients. Show us the recipe." We listened. That request became the foundation of our paper, "Placing the Bust" - and today it becomes a forecast you can open, inspect, and pressure-test yourself.

What's new: the base forecast is on the UI

Every client now has access to the base data center forecast directly in the Vulcan interface, with the full methodology documented alongside it. Nothing is a black box. You can see how we get from satellite-verified online dates to a median outlook, WPI Online Date, and where the model applies land-clear slip, NA delay, and owner-based cancellation trends.

We won't pretend it's the last word. There is still room to build a better forecast, and we'll keep improving it - but this is the best view the available information supports today. For longer-dated projects, we extend our near-term (0–3 year) knowledge and push projects out at a similar rate rather than inventing certainty we don't have.

Figure 1. The new SynMax Vulcan Data Center Base Forecast, now live on the UI — satellite-verified online-date model with median outlook through 2035, viewable for the U.S. and the rest of the world.

Why 2026 looks different from consensus

Here is the part worth slowing down for. Market consensus puts 2026 at roughly 21 GW. Our forecast is materially more grounded - because the binding near-term constraint isn't nameplate ambition, it's the advanced-packaging stage. That single bottleneck makes our 2026 numbers more real and viable than consensus, and it's exactly where a top-down model goes wrong. "Placing the Bust" walks through the detail.

Look further out and the picture flips: in the outer years, capacity projects should be able to satisfy at least the U.S. market. HBM projects are being added to the model, and while they matter, they don't show up as the long-run constraint either. The honest caveat - also in the paper - is that any given year carries a real chance of a supply-chain disruption somewhere in the chain.

Coverage is strongest where our monitoring is deepest. The U.S. forecast is the most fully processed, with 90%+ of projects monitored. The rest of the world is expanding now and should reach U.S.-level coverage within a few months.

Coming next: an Industrial category

We're adding a new category to the UI - Industrial - to house projects like the advanced-packaging campuses featured in the report. It's the same satellite-verified, milestone-by-milestone monitoring you rely on for data centers, pointed at the supply chain that feeds them.

Figure 2. Vulcan construction-timeline monitoring of the TSMC AP7 advanced-packaging campus in Chiayi, Taiwan. Fab 1's eighteen-month gap between land clearing and first structures reflects an on-site archaeological delay; Phase II is tracked to a Nov 2027 earliest / Jun 2028 median online date.

If you're already a client

You should already have an email with access to "Placing the Bust." Open the base forecast on the UI, read the methodology, and put it up against your own view. If it moves your outlook — or if it doesn't — we want to hear why.