- Samsung co-led a €200 million Series A funding round for Euclyd, a Dutch AI chipmaker founded in 2024 that is designing an alternative processor architecture for AI inference workloads. The round also included Somerset Capital Partners, the Scaleup Europe Fund managed by EQT, and Innovation Industries, valuing the startup at roughly $400-500 million.
- Euclyd’s silicon is specifically engineered for inference—the process of running trained AI models in production—rather than training, targeting different economics than Nvidia’s GPUs. The company claims its architecture will significantly reduce energy consumption and operational costs for AI data center infrastructure, areas where hyperscalers are seeking alternatives to Nvidia’s expensive and power-hungry systems.
- The investment reflects a growing trend among tech giants to develop proprietary AI chips: OpenAI announced its Jalapeño chip in August, while Google, AWS, Meta, and others have launched or accelerated their own silicon programs. Euclyd plans to operate two revenue streams—selling hardware and rack systems to enterprise customers for on-premise AI, and licensing its intellectual property to companies building their own chips.
- Samsung’s strategic value extends far beyond capital: the company is one of the world’s largest memory manufacturers and possesses deep expertise in supply chain management, semiconductor design, and systems integration. Euclyd aims to deploy physical chip systems by 2028 and serve thousands of enterprise customers by 2030, though its technology remains unproven at commercial scale.
What Happened?
Samsung led a €200 million Series A funding round for Euclyd, a Dutch AI chip startup developing inference processors designed to compete with Nvidia’s GPUs. The round was co-led by Somerset Capital Partners, the Scaleup Europe Fund managed by EQT, and Innovation Industries. Euclyd, founded in 2024, is designing chip systems with fundamentally different architecture from GPUs, optimized for running trained AI models at scale while reducing energy consumption and operational costs. The company plans to begin deployments in 2028 and serve thousands of enterprise customers by 2030.
Why It Matters?
For investors and technology allocators, this deal signals accelerating fragmentation of the AI semiconductor market away from Nvidia’s near-monopoly. Hyperscalers—Google, AWS, Meta, and now Samsung itself—are racing to develop proprietary chips to reduce both costs and dependence on Nvidia. The investment validates inference as a distinct and massive market opportunity; as models mature, inference workloads will likely dwarf training costs. For equity investors in Nvidia, this represents one of many credible competitive threats emerging in 2026. For enterprise customers, a working alternative could reduce AI infrastructure expenses by 20-40% if Euclyd’s claims about efficiency hold at scale.
What’s Next?
Watch for Euclyd’s first commercial deployments and customer announcements in late 2027 or early 2028. If the startup hits its performance targets, expect other European and Asian semiconductor companies to pursue similar designs. Monitor Nvidia’s response—the company may accelerate software integration or partnerships to lock in customers. Track whether Samsung uses Euclyd chips in its own data centers or cloud offerings, signaling confidence in the technology and creating a reference customer at scale.
Source: CNBC
