What happened
Euclyd, a Netherlands-based semiconductor startup, has raised $231 million in a Series A funding round co-led by Somerset Capital Partners, the Scaleup Europe Fund, and Innovation Industries, with participation from Samsung. The company is developing a proprietary processor and memory architecture designed specifically for AI workloads. Euclyd plans to offer both physical rack systems for on-premises enterprise use and an intellectual property licensing model for firms looking to build their own custom AI chips. The company does not expect to begin commercial hardware shipments until 2028.
Euclyd, founded in 2024, has secured $231 million in Series A funding to develop an alternative to Nvidia’s GPU-centric AI hardware. The round was co-led by Somerset Capital Partners, the Scaleup Europe Fund, and Innovation Industries, with strategic participation from Samsung.
The startup’s business model is twofold: selling physical AI rack systems for on-premises enterprise deployment and licensing its proprietary processor and memory architecture to other firms. The company claims this approach will allow organizations to manage AI workloads with greater control over their data and infrastructure.
Samsung’s role is described as more than financial; the company is expected to provide engineering support and supply chain expertise, leveraging its position as a global leader in memory production to help Euclyd address the data-transfer demands of modern AI workloads.
Despite the funding, Euclyd has not yet demonstrated its technology at a commercial scale. The company has set a target for commercial hardware shipments to begin in 2028, with a long-term goal of supporting thousands of enterprise clients by 2030.
Source details: rswebsols.com ↗
Why it matters
The investment highlights the ongoing industry-wide effort to move beyond general-purpose GPUs, which currently dominate the AI computing landscape. By focusing on a specialized architecture that integrates processing and memory, Euclyd aims to address the data transfer bottlenecks that limit current AI infrastructure. Samsung’s involvement is particularly significant, as the company’s expertise in memory manufacturing and supply chain management provides Euclyd with a strategic partner to navigate the complex semiconductor development cycle. However, Euclyd faces substantial hurdles, including the need to compete with Nvidia’s established software ecosystem and the technical risks inherent in bringing a new hardware architecture to market. The success of this venture remains speculative, as the company has yet to demonstrate its technology at a commercial scale, and its product roadmap extends several years into the future.
The current AI hardware market is heavily reliant on Nvidia’s GPUs, which were originally designed for gaming but have become the standard for AI training and . This reliance has created significant cost and supply chain pressures for major cloud providers and AI enterprises.
Euclyd’s focus on a specialized architecture that integrates memory and processing is a direct attempt to solve the 'memory wall'—the bottleneck caused by moving large volumes of data between processors and memory. If successful, this could offer higher efficiency for tasks compared to general-purpose hardware.
The competitive landscape is intensifying, with major players like Google, Meta, and OpenAI developing their own proprietary AI chips. Euclyd must prove that its hardware can provide a meaningful performance advantage while also overcoming the 'moat' created by Nvidia’s extensive software and developer ecosystem.
The investment carries significant technical and commercial risk. Because the company is still years away from shipping, it must navigate the volatility of the semiconductor industry while proving its design is viable for mass production.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
What is the best response when AI Models Explained makes a mistake in production?
What to watch next
The primary milestone for Euclyd is the transition from architectural design to commercial production, with shipments targeted for 2028. Observers should monitor whether the company can successfully integrate its hardware with existing AI software frameworks, a critical requirement for enterprise adoption. Additionally, the company’s ability to meet its goal of serving thousands of enterprise clients by 2030 will depend on its performance benchmarks and efficiency metrics compared to the rapidly evolving offerings from established incumbents and other startups developing proprietary AI silicon.
Watch for future technical disclosures regarding the performance and power efficiency of Euclyd’s architecture compared to current industry standards.
Monitor the company’s progress in building a software ecosystem, as hardware performance alone is rarely sufficient to displace established platforms in the enterprise market.
Track the company’s ability to maintain its 2028 shipping timeline, as delays in semiconductor manufacturing are common and could impact the startup’s competitive position.