SPRIND and NADI put €40m behind AI-designed chips, with applications open until 30 November
Germany's breakthrough innovation agency and its new Dutch counterpart will fund seven teams with up to €2.6m each, then three of them with up to €7m each.
Germany's Federal Agency for Disruptive Innovation (SPRIND) and the Netherlands' National Agency for Disruptive Innovation (NADI) announced on 23 September 2026 a joint AI-Native Chip Design Challenge worth €40 million. It is NADI's first programme. The application period runs until 30 November 2026 at 11:00 CET.
The target is the design cycle rather than the fab. SPRIND states the goal as “Designing high-efficiency AI Chips in weeks, not in years, with AI-Native Systems”, and asks teams to produce production-ready chip designs for AI workloads in which AI itself optimises the workflow autonomously, from architecture search through verification. Europe has spent most of the past decade putting public money into manufacturing capacity; this puts it into the step before silicon is cut.
The structure is a two-stage tournament running 20 months in total. In Stage 1, from November 2026 to July 2027 and lasting eight months, seven teams receive up to €2.6 million each. Three of them advance to Stage 2, which runs twelve months and carries up to €7 million per team. That is up to €9.6 million per team across both stages. SPRIND says follow-on funding for a tape-out phase, in the tens of millions of euros, will be announced later. Detailed eligibility sits in the call for submission and the participation agreement rather than on the challenge page.
“The launch of NADI's first programme is a huge statement of Europe's intent to combine two of the continent's greatest strengths: collaboration and innovation,” said Jano Costard, head of challenges at SPRIND. Jelle Prins, co-founder of NADI, said: “NADI's mission is to catalyse breakthrough innovation in areas of significant societal and economic importance.” The Challenge format is SPRIND's own instrument: it funds several competing teams in parallel against a technical goal and cuts the field at defined milestones, rather than picking one winner at the outset.
Why it matters for AgTech founders: read this twice, because the instrument is more portable than the topic. Up to €9.6 million over 20 months with no equity given up is larger than most European AgTech seed and Series A rounds, and it is awarded against a technical milestone rather than traction, pipeline or revenue. That is the opposite of what a venture fund screens for, which makes challenge-type money the cheapest capital available to anyone whose main risk is whether the thing works at all. Three practical notes. First, the staging is real: seven teams become three, so plan runway on the assumption that Stage 2 does not arrive, and treat Stage 1 as funded de-risking rather than a growth round. Second, the deadline is eight weeks out, and consortium agreements are what usually make teams miss these calls, not the technical write-up. Third, the downstream relevance is concrete even if you never apply. What a sprayer, a sorting line or a camera boom can afford to run per plant is set by the cost and power budget of edge inference silicon, and a shorter European design cycle is what would eventually make application-specific ag hardware something other than a thought experiment. SPRIND runs challenges in agriculture and biology too; the agency is worth tracking on its own terms.
SPRIND and NADI put €40m behind AI-designed chips, with applications open until 30 November
Germany's breakthrough innovation agency and its new Dutch counterpart will fund seven teams with up to €2.6m each, then three of them with up to €7m each. The money is non-dilutive.
Germany's Federal Agency for Disruptive Innovation (SPRIND) and the Netherlands' National Agency for Disruptive Innovation (NADI) announced on 23 September 2026 a joint AI-Native Chip Design Challenge worth €40 million. It is NADI's first programme. The application period runs until 30 November 2026 at 11:00 CET.
The target is the design cycle rather than the fab. SPRIND states the goal as “Designing high-efficiency AI Chips in weeks, not in years, with AI-Native Systems”, and asks teams to produce production-ready chip designs for AI workloads in which AI itself optimises the workflow autonomously, from architecture search through verification. Europe has spent most of the past decade putting public money into manufacturing capacity; this puts it into the step before silicon is cut.
The structure is a two-stage tournament running 20 months in total. In Stage 1, from November 2026 to July 2027 and lasting eight months, seven teams receive up to €2.6 million each. Three of them advance to Stage 2, which runs twelve months and carries up to €7 million per team. That is up to €9.6 million per team across both stages. SPRIND says follow-on funding for a tape-out phase, in the tens of millions of euros, will be announced later. Detailed eligibility sits in the call for submission and the participation agreement rather than on the challenge page.
“The launch of NADI's first programme is a huge statement of Europe's intent to combine two of the continent's greatest strengths: collaboration and innovation,” said Jano Costard, head of challenges at SPRIND. Jelle Prins, co-founder of NADI, said: “NADI's mission is to catalyse breakthrough innovation in areas of significant societal and economic importance.” The Challenge format is SPRIND's own instrument: it funds several competing teams in parallel against a technical goal and cuts the field at defined milestones, rather than picking one winner at the outset.
Why it matters for AgTech founders: read this twice, because the instrument is more portable than the topic. Up to €9.6 million over 20 months with no equity given up is larger than most European AgTech seed and Series A rounds, and it is awarded against a technical milestone rather than traction, pipeline or revenue. That is the opposite of what a venture fund screens for, which makes challenge-type money the cheapest capital available to anyone whose main risk is whether the thing works at all. Three practical notes. First, the staging is real: seven teams become three, so plan runway on the assumption that Stage 2 does not arrive, and treat Stage 1 as funded de-risking rather than a growth round. Second, the deadline is eight weeks out, and consortium agreements are what usually make teams miss these calls, not the technical write-up. Third, the downstream relevance is concrete even if you never apply. What a sprayer, a sorting line or a camera boom can afford to run per plant is set by the cost and power budget of edge inference silicon, and a shorter European design cycle is what would eventually make application-specific ag hardware something other than a thought experiment. SPRIND runs challenges in agriculture and biology too; the agency is worth tracking on its own terms.
