Samsung has co-led a 200-million-euro (roughly $231 million) Series A round in Euclyd, a Dutch startup designing AI inference chips with an architecture that departs from the GPUs that made Nvidia the world's most valuable company.
Euclyd CEO Bernardo Kastrup told CNBC exclusively that the round was co-led by Samsung alongside Somerset Capital Partners, the Scaleup Europe Fund managed by EQT, and Innovation Industries. The raise is more than double the at least $100 million the company said it was seeking when it spoke to CNBC in April 2026, a sign that investor interest in GPU alternatives has accelerated sharply over the past six months. For more context on this story, see our ongoing artificial intelligence updates.
Why Investors Are Funding GPU Alternatives
Nvidia built a near-monopoly over the market for the highest-end AI processors after its gaming GPUs were repurposed for training and running AI models, work broadly known as inference. That dominance has made chip diversification one of the most active corners of AI investing, with hyperscalers and startups alike working on custom silicon.
OpenAI announced in August that its first in-house AI chip, the Jalapeño, had achieved what the company described as industry-leading speed and efficiency. Google, AWS and Meta are all developing their own AI processors as well, largely to reduce their dependence on Nvidia's pricing and supply constraints.
Euclyd, founded in 2024, is betting that inference workloads in particular are ripe for a redesign. According to CNBC, the company is building an AI chip system with a fundamentally different architecture from GPUs, covering both the processor and the memory architecture that feeds it.
"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 said in the CNBC interview.
A Different Bet: Sell Systems and Licenses, Not GPUs
Rather than chasing the same customers as Nvidia with a lookalike product, Euclyd is targeting two distinct revenue streams. The first is selling hardware and physical rack systems to enterprise customers that want secure, self-hosted AI inference. The second is licensing its intellectual property to companies that want to build their own chips on top of Euclyd's designs.
That dual model reflects a split in the AI infrastructure market between organizations that want turnkey capacity and those, particularly large technology companies, that would rather own the underlying design. It also gives Euclyd a path to revenue before its own systems reach the market.
What Samsung Brings Beyond Capital
Kastrup told CNBC that Samsung's involvement extends well past the check. "Samsung can help us in more ways than money," he said. "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."
The strategic logic is notable given that memory, not just logic, has become a bottleneck for AI data centers. Chipmakers across the industry have struggled to secure enough high-bandwidth memory, and Nvidia has raised AI server prices in part because of memory costs. A close relationship with one of the world's largest memory manufacturers could give Euclyd an advantage in co-designing processor and memory as a single system.
The Catch: Nothing Is Proven at Scale Yet
For all the momentum, Euclyd's systems have yet to be proven in commercial deployments at scale, as CNBC noted. The company says its silicon for foundational models will cut the energy needs and costs of AI data center infrastructure, but those claims will not be tested in earnest until hardware ships.
Kastrup told CNBC that Euclyd aims to begin rolling out its physical chip systems in 2028, with the goal of serving thousands of enterprise customers by 2030. That timeline puts the company's commercial debut years behind Nvidia's current product cycle, meaning it will be betting against a moving target.
Europe's Growing Chip Ambitions
The round also underscores Europe's push to claim a slice of the AI hardware stack. The Scaleup Europe Fund is managed by Swedish investment firm EQT, and Innovation Industries is a Dutch deep-tech investor. Combined with Samsung's global manufacturing reach, the investor mix gives the Eindhoven-area startup an unusually international base for a company founded just two years ago.
Euclyd is part of a wider wave of well-funded Nvidia challengers that has attracted record levels of venture capital as competition in AI silicon intensifies. Whether any of them can dent Nvidia's lead remains an open question, but the size of Samsung's commitment suggests the world's largest chipmakers are no longer willing to treat the matter as hypothetical.
For enterprises planning AI infrastructure investments toward the end of the decade, the practical takeaway is that the inference market they will be buying from in 2028 and beyond is likely to look considerably more crowded than it does today.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
