- Euclyd raised over €200 million in a Series A led by Samsung and three funds.
- Ex-ASML chief Peter Wennink joins as chairman, doubling the round Euclyd first sought.
- The Dutch startup is building AI inference chips to challenge Nvidia’s dominance.
Bernardo Kastrup drew up Euclyd’s first designs in his attic, working at night after his day job at ASML ended. Two years on, Samsung and the man who used to run ASML are the ones writing the cheques.
Euclyd, the Eindhoven-based chip startup, has closed a Series A of more than €200 million, co-led by Samsung, Somerset Capital Partners, the Scaleup Europe Fund managed by EQT, and Innovation Industries. EIFO, imec.xpand, the Brabant Development Agency, and Quadri also joined.
From five founders to a full-stack chip company
Euclyd was founded in 2024 at Eindhoven’s High Tech Campus by Bernardo Kastrup, Atul Sinha, Gerard Egelmeers, Ingolf Held and Harm Peter.
Kastrup, who holds PhDs in both philosophy and computer engineering, previously co-founded Silicon Hive, a parallel-processor company Intel acquired in 2011, and worked as a technology strategist at ASML.
Euclyd’s pitch is that GPUs are the wrong tool for AI inference—the stage where a trained model actually answers a prompt. GPUs spend most of their energy moving data between memory and compute cores.
Its chip, CRAFTWERK, packs 16,384 custom processors that work directly on data in memory. A rack-scale version, CRAFTWERK Station, combines 32 of those chips and targets one exaflop of compute by 2028.
The company claims up to 100 times the power efficiency of Nvidia’s latest Vera Rubin chips, modelled on Meta’s Llama 4 Maverick. That’s a company-supplied projection, not a result from a live deployment, and it hasn’t been tested at commercial scale.
The race beyond NVIDIA
Euclyd isn’t the only one chasing Nvidia’s inference gap. Axelera AI, another Eindhoven company, raised more than $250 million in February for edge inference.Etched’s valuation rose from $5 billion in June to $21 billion by August, on the back of a working chip and a Jane Street customer contract. OXMIQ, which Tech Funding News covered in July, is skipping chip manufacturing altogether and instead licensing its architecture.
Where Axelera is chasing the edge and OXMIQ is chasing licensing revenue, Euclyd is trying to own the whole stack: chip, memory and the datacenter systems around them. It’s a more expensive bet, and a slower one to prove out.
Samsung’s Dede Goldschmidt, senior vice president and head of its Semiconductor Innovation Center, put it this way: “The next phase of AI will be defined not only by model innovation, but also by the efficiency and scalability of its infrastructure.”
Kastrup added the round “accelerates our mission to make intelligence abundant through greater efficiency, lower cost per token, and broader access to advanced AI.”
The number has already doubled once
Euclyd told CNBC in April it was seeking at least €100 million. By June, Bloomberg had it in talks for up to €200 million. The final number cleared that ceiling, on top of a seed round of under €10 million.
The new capital will grow Euclyd’s engineering team, speed up its silicon and systems roadmap, and prepare it for enterprise, sovereign and hyperscale customers. The company says it’s in talks with four prospective customers and hopes to start supplying two of them in 2027.
The global AI chip market is projected to grow from roughly $100 billion this year to more than $400 billion by 2030, and Nvidia still controls an estimated 80% of AI training chip demand.
Whether that thesis survives contact with actual production at a two-year-old company with no chips yet shipping is the question the next 18 months will have to answer.
