Netherlands and Germany Launch €40 Million Partnership for Energy-Efficient AI Chips

2026-09-23 semicon

The Hague, Wednesday 23 September 2026
The Dutch agency NADI and Germany’s SPRIND have launched a €40 million challenge to develop energy-efficient AI chips, aiming to reduce Europe’s reliance on foreign technology.

A Sovereign Spark in the Dutch Ecosystem

The newly established Dutch National Agency for Disruptive Innovation (NADI) was officially announced on Prinsjesdag, 15 September 2026, by Minister of Economic Affairs Heleen Herbert, backed by a €500 million government fund in the national budget [3][8]. Modelled after the United States’ Defense Advanced Research Projects Agency (DARPA), NADI is designed to fund high-risk, high-reward technological breakthroughs that are often deemed too volatile for traditional venture capital firms [1][4][6]. The launch of the “AI-Native Chip Design Challenge” on 22 September 2026 represents NADI’s first major bilateral initiative since its inception this month, signaling a highly targeted, sovereign-backed approach to high-tech funding [2][3][8].

Collaborating Across the Rhine

By partnering with Germany’s Federal Agency for Disruptive Innovation (SPRIND), which has been operating since 2019 [3][8], NADI is leveraging an established, highly successful model. SPRIND has historically generated an average of €2.80 in private follow-on financing for every €1.00 of public investment [3]. This joint €40 million programme seeks to address a critical bottleneck in the European semiconductor industry: the early-stage funding gap in chip design [3][8]. As Beau-Anne Chilla, co-founder of NADI, points out, Europe possesses one of the strongest chip ecosystems globally, yet its startups struggle to secure capital during the high-risk initial design phase, a gap that NADI intends to fill [3][8].

Accelerating Design from Years to Weeks

The core mission of the AI-Native Chip Design Challenge is to radically compress the chip development timeline from several years to just a few weeks [2][8]. Traditional chip design is a notoriously slow, multi-layered process [7]. By utilizing artificial intelligence to optimize and guide the design process, NADI and SPRIND hope to achieve what SPRIND’s Head of Challenges, Jano Costard, describes as a “several orders of magnitude acceleration” [1]. The initiative targets the development of energy-efficient chips optimized for both the training and inference phases of AI models [1][4].

Addressing the Energy Crisis and “Grocery Trucks”

The focus on energy efficiency is a direct response to the soaring power demands of modern AI infrastructure [6]. Jelle Prins, co-founder of NADI, colourfully illustrated the current inefficiency by comparing the use of standard Nvidia chips for everyday tasks to “doing your shopping with a removal truck” [6][7]. To solve this, the challenge’s application window opens on 23 September 2026 and closes on 30 November 2026 [3][8]. The program will select seven initial teams, with each receiving €2.6 million—representing a collective first-phase commitment of €18.2 million [3][6][8]. After an eight-month development phase, the three most promising teams will be selected to receive a further €7 million each, representing an additional €21 million to bring their designs closer to commercialisation [3][6][8].

Securing Europe’s Technological Sovereignty

This bilateral initiative comes at a critical juncture for European strategic autonomy. Currently, European AI models run almost exclusively on foreign-designed chips, leaving the continent highly dependent on American and Asian supply chains [3][6][8]. This effort aligns with broader European initiatives, such as the European Union’s commitment on 31 July 2026 to allocate €5 billion towards funding seven AI “megafactories” to compete with the US and China [1]. By combining the Netherlands’ world-class photolithography and semiconductor equipment ecosystem—anchored by ASML—with Germany’s robust industrial manufacturing and research capabilities, the alliance aims to build a resilient, self-sustaining semiconductor value chain [1][4][6].

Sources & Ecosystem Partners

Innovation policy AI chip design