Samsung Electronics has officially thrown its weight behind Euclyd, a fledgling semiconductor startup based in the Netherlands that is developing alternative hardware designed to challenge Nvidia’s near-monopoly on artificial intelligence infrastructure. Euclyd announced the successful close of a $231 million (€200 million) Series A funding round, with the South Korean tech conglomerate participating alongside a prominent consortium of European investors, including Somerset Capital Partners, the Scaleup Europe Fund, and Innovation Industries.
Despite the substantial capital injection, commercial hardware from the Dutch enterprise will not reach the market until 2028, according to Euclyd Chief Executive Bernardo Kastrup. The lengthy development timeline underscores the immense engineering hurdles involved in designing next-generation silicon capable of competing with the entrenched architectures of Silicon Valley giants. Nevertheless, the backing of a tier-one memory manufacturer like Samsung signals that venture capitalists and industry incumbents alike are aggressively hunting for alternatives to the current hardware paradigm governing the artificial intelligence boom.
A Two-Pronged Commercial Strategy
Founded in 2024, Euclyd enters a semiconductor landscape defined by explosive demand and severe supply chain bottlenecks. In an exclusive interview following the funding announcement, Kastrup outlined a two-pronged business model designed to maximize the startup’s market penetration over the coming decade.
The first revenue stream involves selling physical hardware and integrated rack systems directly to enterprise clients. These systems are aimed at organizations seeking to run complex AI inference workloads securely and efficiently on-site, bypassing the need to rely entirely on public cloud providers. The second approach focuses on intellectual property (IP) licensing, allowing external firms to incorporate Euclyd’s proprietary chip designs into their own custom silicon manufacturing pipelines.
Unlike conventional graphics processing units (GPUs) originally built for video rendering that have since been repurposed for AI training and inference, Euclyd is engineering a radically different architecture. The startup’s system architecture re-imagines the integration between processor and memory layers from the ground up.
"AI is becoming a foundation of economic growth, scientific discovery and national competitiveness, but its potential will remain constrained unless we fundamentally change the infrastructure beneath it," Kastrup stated.
Strategic Value Beyond Capital
While the €200 million cash infusion provides Euclyd with a robust runway, Kastrup emphasized that Samsung’s participation brings invaluable strategic advantages that far outweigh financial investment. As one of the world’s largest manufacturers of memory chips—a critical component in high-performance AI accelerators—Samsung offers deep supply chain visibility, sophisticated manufacturing expertise, and an extensive global enterprise network.
"Samsung can help us in more ways than money," Kastrup noted. "They are one of the biggest memory manufacturers in the world. They do a lot of engineering, they know a lot about systems, they know the supply chain, they have a huge network."
This partnership addresses one of the primary vulnerabilities faced by hardware startups: the transition from theoretical chip design to scalable physical production. Semiconductor fabrication requires access to cutting-edge cleanrooms, lithography equipment, and advanced packaging technologies—resources typically concentrated among a handful of dominant players like TSMC and Samsung. By securing an ally in Samsung, Euclyd gains a potential manufacturing pathway that could otherwise take decades to establish independently.
The Broader Landscape: Escaping the Nvidia Hegemony
Euclyd’s emergence arrives at a time of intense industry-wide anxiety over hardware availability and skyrocketing costs. Nvidia’s ascent to becoming one of the world’s most valuable companies has been fueled by its dominance in AI training and high-end inference chips. However, the heavy financial burden of purchasing and maintaining Nvidia hardware has prompted technology leaders to seek alternatives.
The movement toward custom silicon is accelerating across the technology sector. In August, OpenAI revealed that its inaugural in-house artificial intelligence chip, internally designated as Jalapeño, delivered industry-leading speed and efficiency metrics during preliminary testing. Meanwhile, hyperscale cloud providers such as Google, Amazon Web Services (AWS), and Meta are heavily investing in proprietary custom chips—such as Google’s Tensor Processing Units (TPUs) and AWS’s Trainium and Inferentia lines—to handle internal workloads and reduce their reliance on merchant silicon providers.
Despite these industry shifts, market analysts urge caution regarding the realistic prospects of early-stage startups. Euclyd’s systems remain entirely unproven at any meaningful commercial scale. Transitioning from a 2024 startup foundation to shipping silicon by 2028, with an ambitious target of serving thousands of enterprise customers by 2030, leaves little room for operational missteps.
Timeline and Future Outlook
The trajectory of Euclyd over the next several years highlights the grueling pace of modern semiconductor development:
- 2024: Euclyd is founded in the Netherlands, initiating fundamental research into alternative AI processor and memory architectures.
- Late 2026: The company successfully closes its Series A funding round, securing $231 million (€200 million) co-led by Samsung, Somerset Capital Partners, the Scaleup Europe Fund, and Innovation Industries.
- 2028: Projected window for the commercial release and initial shipment of Euclyd’s enterprise hardware and rack systems.
- 2030: Corporate milestone target to scale operations and serve thousands of enterprise customers worldwide through direct hardware sales and IP licensing agreements.
Whether a small European startup with no commercially shipped products can meaningfully chip away at Nvidia’s market stronghold remains an open question. The barriers to entry in the semiconductor industry are historically high, requiring billions of dollars in continuous research and development, software ecosystem maturity, and developer buy-in.
Nvidia’s competitive moat is protected not just by raw hardware performance, but by CUDA, its proprietary software ecosystem that has become the industry standard for AI development. For Euclyd and other challengers to succeed, hardware innovations must be matched by software compatibility that makes migration seamless for developers.
As venture capital continues to flow into alternative silicon designers, the semiconductor ecosystem is steadily decentralizing. While Nvidia remains firmly entrenched at the apex of the artificial intelligence revolution, the collective efforts of startups like Euclyd, alongside the in-house hardware initiatives of tech conglomerates, point toward a more diversified and competitive future for global AI infrastructure.
