
NADI and SPRIND launch a €40M challenge to accelerate AI chip design
The Netherlands' new DARPA-style agency and Germany's SPRIND will commit €40 million over 20 months to small teams using AI to cut chip design times from years to weeks.
The Dutch National Agency for Disruptive Innovation (NADI) and Germany's Federal Agency for Breakthrough Innovation (SPRIND) said on Wednesday they will commit €40 million over 20 months to groups of small teams using artificial intelligence to accelerate the design of chips for training and running AI models. It is the first joint funding programme between the two agencies and the first challenge run by NADI.
A Dutch answer to DARPA
NADI was presented on 15 September 2026 by Dutch economic affairs minister Heleen Herbert on Prinsjesdag, the country's annual budget day. The government has reserved €500 million (about $570 million) for the agency, which is modelled on the US Defense Advanced Research Projects Agency — the 1958 body behind technologies that led to the internet, stealth aircraft and GPS — as well as on SPRIND, which has worked that way since 2019, and on Britain's ARIA. Entrepreneur Jelle Prins, NADI's co-founder, is building up the organisation alongside deeptech investor Beau-Anne Chilla, after a report led by former ASML chief executive Peter Wennink recommended a deployment budget of €1.5–2 billion.
Seven teams, then three
The AI-Native Chip Design Challenge runs in two stages. In the first, from November to July 2027, seven teams will each receive €2.6 million; applications are open from 23 September to 30 November. A second stage lasting a further year will shortlist three teams for €7 million each. SPRIND provides the majority of the funding and NADI the remainder, and participating teams also receive technical and commercial coaching.
From years to weeks
The aim is to compress a design process that today takes years into weeks. The agencies point to AI models that can now draw technical conclusions and to reinforcement-learning agents that independently solve complex optimisation problems, with the longer-term prospect of AI handling chip design autonomously. The designs are intended for high-performance AI training and inference, with better energy efficiency and lower inference costs.
Prins said a partnership with Germany on chips was natural, given the ecosystem around equipment maker ASML in the Netherlands and German strengths in research and manufacturing. Jano Costard, SPRIND's head of challenges, told Reuters the goal was "several orders of magnitude acceleration". Europe currently depends on the US and China for semiconductor supply. SPRIND is separately running a €125 million challenge to create three European frontier-AI labs, and says its challenge model has committed more than €100 million in 2026.
SiTech — AI-powered web development
We build fast, modern websites and bring AI into real business workflows. Have a project or a question? We'd love to help.