Jacob Fehn
From the grid to the GPU.
- Networks, power and buildings for Nordic AI data centers
- Author of The Definitive Guide to AI Data Centers
- Maps grid capacity for siting with Substralis
- Based in Oslo
Talks
Saw one of these live? The code on the last slide unlocks the slides as a PDF.
Work
The Definitive Guide to AI Data Centers
A vendor-neutral engineering and economics atlas for siting, building, financing, commissioning and operating AI data centers.
aidatacenterguide.comSubstralis
The grid capacity Nordic and UK operators publish, on one map: station capacity, connection queues and outlooks from official registers.
substralis.com
About
I work with the people building AI data centers in the Nordics: the operators, the neoclouds and the engineers who have to make a GPU cluster actually run.
A site needs power, cooling, a building, a network and financing before it can serve a single model, and the early decisions in each of those limit the others for decades. Most of what I do is explain how those pieces fit together, in talks, in writing and in tools.
Short bio for event organisers
Jacob Fehn works with the companies building AI data centers in the Nordics and is based in Oslo. He is the author of The Definitive Guide to AI Data Centers (aidatacenterguide.com) and builds Substralis, a map of the grid capacity Nordic and UK operators publish.
Contact
Questions about a talk, or an invitation to give one:
jacob@fehn.ai