The startups fixing Europe’s AI power problem
Earlier this year, I wrote about the AI power crisis and argued that coordination now matters more than new capacity. The numbers have not improved. Research from the International Energy Agency (IEA) shows that global data centre electricity demand is projected both to roughly double, from 485 TWh in 2025 to around 950 TWh by […] The post The startups fixing Europe’s AI power problem appeared…
European startups are addressing the AI power crisis by focusing on software solutions that optimize existing infrastructure. The International Energy Agency (IEA) predicts that global data center electricity demand will double by 2030 and grow more than four times faster than overall electricity demand. In Europe, a small but growing group of companies is working on four efficiency layers: grid, facility, compute, and software.
Grid efficiency companies like Amsterdam's Sympower, Dublin-based GridBeyond, Munich's Entrix, Finland's Capalo AI, London's Piclo, and Vienna's enspired make existing capacity visible by aggregating demand and bidding it into balancing markets. Facility optimization company etalytics optimizes cooling, heating, and power systems inside facilities, reporting savings up to 50%.
Rotterdam's Gradyent optimizes district heating networks for energy companies across 35 European cities. Compute efficiency company FlexAI places AI training and inference workload across heterogeneous hardware, while YellowDog schedules workloads across on-premises, hybrid, and multi-cloud fleets. Software efficiency company Multiverse Computing compresses large language models by up to 95%, and Pruna AI combines pruning, quantization, and other compression methods to make models cheaper to run.
Despite the growing importance of the software layer, no European company currently connects multiple layers to make grid, facility, and compute decisions together. Founders see this as the clearest and most defensible opportunity in AI infrastructure.
Written by urgent.news from EU-Startups's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
