EUCLYD, an AI infrastructure company based in Eindhoven, the Netherlands, has signed a Series A financing round of more than €200 million, the company announced on September 15, 2026. The round was co-led by Samsung, Somerset Capital Partners, the Scaleup Europe Fund managed by EQT, and Innovation Industries. The size of the round and the composition of its investors matter because they signal that corporate and institutional capital is prepared to fund European semiconductor development for AI, not only software and applications.
Who stands behind the company
Beyond the four co-leads, the round includes EIFO, the Export and Investment Fund of Denmark, along with imec. xpand, Brabant Development Agency (BOM), and Quadri. Peter Wennink, former President and CEO of ASML, has joined EUCLYD as Chairman of the Board. The company was founded by Bernardo Kastrup and Atul Sinha at High Tech Campus Eindhoven, which EUCLYD calls Europe's smartest square kilometer, and it states that it draws on the campus's semiconductor ecosystem and Europe's engineering heritage. The new capital is earmarked for growing the engineering organization, speeding up the silicon and systems roadmap, and deepening ecosystem partnerships, while readying the company for commercial deployment in enterprise, sovereign, and hyperscale AI markets.
The company frames the problem it targets as an efficiency wall. In its description, each gain in foundation-model capability demands more power, more memory bandwidth, and more capital to deploy, while the underlying infrastructure grows more complex. EUCLYD's answer is a platform spanning crafted compute, memory architecture, and datacenter systems, intended to break that wall and change the economics of AI inference. At the center of the roadmap sit craftwerk, which the company calls the world's first agentic AI silicon, and craftwerk station CWS, which it calls the world's lowest-power exascale AI factory. On EUCLYD's website the station carries the designation CWS 32.
The design pairs programmable ASIC compute with processor-memory co-design and system-level optimization. EUCLYD presents this combination as a way past the memory and efficiency walls that limit AI performance and scalability. The company's stated long-term vision is Abundant Intelligence, defined as a future where infrastructure cost, power availability, and geography no longer constrain advanced AI. Its homepage presents the same ambition under the banner of AI infrastructure for Abundant Intelligence and frames the goal as bringing inference everywhere, for everyone. The company is also expanding its team across the full compute stack as it enters its next phase of growth, with hiring spanning silicon architecture, design and verification, physical implementation, systems, firmware, compilers, and runtime software.
What this means for business
For companies that run or buy AI inference, the practical question is cost per token and access to capacity. Kastrup, EUCLYD's founder and CEO, said the financing speeds the company's mission to make intelligence abundant, citing greater efficiency, lower cost per token, and broader access to advanced AI. If the roadmap holds, buyers of AI compute gain an additional vendor option in a market where capacity and power availability often determine what a company can deploy. The effect differs by size: a large enterprise or hyperscaler can test a new silicon platform inside existing procurement and engineering processes, while a small company will more likely meet such infrastructure indirectly, through cloud providers that adopt it.
What the news does not mean is that the technology is already available for purchase. EUCLYD describes a roadmap and a platform, and the round funds engineering and preparation for commercial deployment, not a shipping product. The claims attached to craftwerk and CWS 32, including the world's first agentic AI silicon and the world's lowest-power exascale AI factory, are the company's own characterizations and require independent verification. When evaluating such a vendor, the questions to ask concern delivery timelines, benchmark methodology, software support for existing frameworks, and the migration path from current accelerators. Dede Goldschmidt, Senior Vice President of Samsung Electronics and Head of the Samsung Semiconductor Innovation Center, said model innovation alone will not define the next phase of AI, pointing also to the efficiency and scalability of the infrastructure beneath it.
The marker to watch is the composition of EUCLYD's engineering team and the first named deployment partners. The company says it is hiring across silicon architecture, design and verification, physical implementation, systems, firmware, compilers, and runtime software, and is looking for engineers prepared to rebuild AI infrastructure from the logic gate up. If those roles are filled and enterprise, sovereign, or hyperscale customers are named within the coming quarters, the European AI silicon thesis moves from financing to execution. Ted Persson, Partner at EQT and Co-Head of the Scaleup Europe Fund, said Europe's pool of engineering talent can produce globally significant technology companies, and Wennink said Europe possesses world-class capabilities across semiconductors, advanced manufacturing, and systems engineering.
