.png)
Methodology: Every two weeks we collect most relevant posts on LinkedIn for selected topics and create an overall summary only based on these posts. If you´re interested in the single posts behind, you can find them here: https://linktr.ee/thomasallgeyer. Have a great read!
Thanks to Pavel Khamiakou, Anita Schüttler, Maike Conrads, Nina Benoit, Alyson Freeman, Lilibeth Bustos Linares, Fernanda A., ST. Shailesh Tyagi, Navveen Balani, Sigrid Braun, Miguel Fernandez, Micha Schildmann, Richard Tarboton, Nick McDonald, Šimon Rákosník, Kevin Leslie, Esmeralda Colombo, Robbie Epsom and Aydin Mir Mohammadi, and everyone else who contributed insights to this edition. Find the full list of posts and voices on LinkedIn: https://www.linkedin.com/pulse/best-linkedin-cw-38-39-green-ict-sustainable-ai-thomas-allgeyer-c4e7e?trackingId=XE2GCao5Ooa6kv4K3Koc5w%3D%3D&lipi=urn%3Ali%3Apage%3Ad_flagship3_profile_view_base_recent_activity_content_view%3B24xBJn%2FeQEyLJEpL%2FDpK1w%3D%3D
With more precise units. Efficiency is measured per business transaction rather than per endpoint, and PUE is giving way to WUE, TUE and watts per token. Smaller models are not always more efficient once retries count, and token billing could raise LLM costs tenfold.
Ratings and impact assessments. The European Commission adopted A to G data centre ratings, and Scotland mandates impact assessments for data centres above 50MW. Transparency gaps remain: AWS's Senate submission omits absolute Australian energy and water use.
A lot at scale. Cutting PUE from 1.54 to 1.09 at 1 GW saves about 3,940 GWh a year. A Viridien immersion retrofit delivers 4x compute on less power, and IBM Diamondback tape cuts storage costs up to 84% and energy 97%.
How we solve real problems for real clients