Institute of Thinking№ 003 — 5 OCT 2026
Accepted to the EuroHPC AI Factories Playground
On 5 October 2026, Institute of Thinking was accepted into the EuroHPC Joint Undertaking’s AI Factories Playground, which gives AI companies access to Europe’s public supercomputing infrastructure.
Our allocation is on Leonardo Booster, the GPU module of Leonardo, the supercomputer operated by CINECA in Bologna, Italy. The Booster module runs 13,824 NVIDIA A100 GPUs across 3,456 nodes. Leonardo entered the TOP500 list in November 2022 as the fourth fastest system in the world and the second in Europe (Wikipedia, Leonardo).
VIDEO: LEONARDO SYSTEM DESIGN — LEONARDO PRE-EXASCALE SUPERCOMPUTER · LEONARDO WEBSITE
What it means for Devotion
Leonardo Booster is where we will train the next versions of Devotion, our multilingual full-duplex voice API. Supercomputer-scale compute lets us run larger training runs across more languages, in less time.
That moves our research forward and shortens the path to commercial applications. Revenue from those applications funds the next round of research.
Our thanks to the EuroHPC Joint Undertaking and CINECA.
