The development of modern semiconductors is complex, costly, and time-consuming. While AI models evolve in short innovation cycles, the development of the necessary hardware often takes several years.

With the challenge "AI-Native Chip Design" the Federal Agency for Disruptive Innovations SPRIND and the Dutch innovation agency NADI, which is currently being established, aim to fundamentally change this process.

The goal is not only to use artificial intelligence as a supportive tool, but to establish it as a central component of chip development.

From searching for suitable chip architectures to RTL generation (Register Transfer Level) to verification, AI systems are intended to largely take over complex development tasks independently.

In particular, reinforcement learning and autonomous AI agentswill be used to evaluate different design options, make decisions, and continuously optimize development processes.

The vision: Powerful chips should be designed in a few weeks instead of over several years.

New chip architectures for AI applications

The focus is on novel, energy-efficient chip architectures that are specifically developed for the requirements of modern AI systems.

Solutions are sought that enable significant advancements in computing power, energy efficiency, and scalability compared to existing approaches.

Particular focus is on the requirements of future AI gigafactories as well as on European technology fields such as data security, biotechnology, robotics, and mobility.

The challenge explicitly does not aim at incremental improvements to existing chip designs. Instead, fundamentally new approaches are to be developed that have the potential to change the development of specialized AI hardware.

In the end, concrete, production-ready chip designs should be available, which can serve as a basis for subsequent manufacturing.

Up to 9.6 million euros per team

The challenge extends over 20 months and two competitive stages. An independent expert jury supports SPRIND and NADI in selecting and evaluating the teams.

Financial support is provided through individual fixed-price contracts based on the respective cost estimates.

In addition, SPRIND and NADI support the participating teams in accessing experts and potential cooperation partners.

For particularly successful teams, follow-up financing in the double-digit million range is also sought after the second stage. This should enable the implementation of the developed designs in the so-called tape-out phase, where a finished chip design is prepared for semiconductor manufacturing.

However, follow-up financing has not yet been promised. Further details will be published at a later date.

Application deadline

Applications for the "AI-Native Chip Design" challenge are possible until November 30, 2026, at 11:00 AM (CET).

For interested teams, SPRIND is hosting an informational webinar on November 17, 2026, from 4:00 to 5:30 PM about the challenge.

The call for applications is aimed at teams pursuing new approaches for a largely AI-driven development of powerful semiconductors.

Further information on the participation conditions and application can be found on the official challenge page of SPRIND.

Impulses for Europe's AI hardware

The challenge connects two central innovation fields: artificial intelligence and semiconductor development.

While the performance of AI applications increasingly depends on available computing capacity and energy efficiency, the use of autonomous AI systems could significantly accelerate the development of specialized hardware.

For the European AI ecosystem, this opens up the prospect of adapting new chip architectures more quickly to technological requirements and expanding its own competences in hardware design.

Whether development times can really be reduced from years to weeks is something the challenge is intended to demonstrate in practice.