Dutch AI Chip Startup Euclyd Raises €200 Million to Challenge Nvidia

Dutch AI chip startup Euclyd has raised more than €200 million (approximately $231 million) in a Series A funding round. The investment marks a significant European effort to develop high-efficiency hardware capable of challenging Nvidia's dominance in the AI inference market.

The funding round was co-led by a consortium including Samsung, Somerset Capital Partners, Innovation Industries, and the Scaleup Europe Fund, which is managed by EQT. Alongside the capital injection, Euclyd has strengthened its leadership by appointing Peter Wennink, the former president and CEO of ASML, as chairman of the board.

The Architecture of Efficiency

Founded in 2024 in Eindhoven by former Philips Research scientist and ASML product strategist Bernardo Kastrup, along with Atul Sinha, Euclyd is positioning itself as a strategic alternative to traditional GPUs. The company's proprietary "Craftwerk" architecture utilizes 16,384 custom SIMD processors designed to operate directly on data stored in memory. This approach is intended to drastically reduce the energy consumption typically associated with moving data between memory and the processor, a primary bottleneck in large-scale AI deployments.

A Push for Technological Sovereignty

This investment represents a broader strategic push for European "technological sovereignty" in the AI sector. By combining the expertise of ASML alumni with the manufacturing and supply chain capabilities of Samsung, Euclyd seeks to reduce the industry's heavy reliance on U.S.-based hardware for the inference stage—the process where trained AI models are actually deployed to generate results for users.

CEO Bernardo Kastrup noted that the partnership with Samsung provides value beyond capital, stating that "Samsung can help us in more ways than money." This collaboration is critical as Euclyd enters a crowded field of inference competitors, including startups like Groq, Cerebras, and Axelera AI, as well as internal silicon projects from tech giants such as Meta and Google.

Market Implications

For the broader AI industry, the emergence of specialized inference chips is a response to the unsustainable power requirements of current GPU clusters. While Nvidia continues to lead in the training of massive models, the cost and energy overhead of running those models at scale have created a market opening for architectures that prioritize power efficiency over general-purpose flexibility.

What Remains to be Seen

While the funding and leadership appointments are confirmed, the company's ability to translate its architecture into mass-produced silicon remains the primary hurdle. The industry will be watching for verified performance benchmarks and the successful transition from design to physical hardware rollouts to determine if Euclyd can realistically displace the current hardware incumbents.