Samsung Just Bet $230M Against Nvidia’s Chips

A four-year-old startup out of Eindhoven just pulled in more than 230 million dollars to build a chip that has nothing to do with training AI models. It’s aimed at the other half of the job, inference, which is the part where a trained model actually answers your question.

The company is called Euclyd, and the round was co-led by Samsung, alongside a few European investors. That alone tells you something. Samsung makes memory and chips for a living, and it just put real money behind the idea that GPUs, Nvidia’s bread and butter, aren’t actually the right tool for this job.

Their pitch is straightforward. Running inference at scale on GPUs is expensive and wastes power, because GPUs were built for training, not for answering millions of prompts a day as cheaply as possible. Euclyd is designing chips and memory architecture built only for that second job.

They’re not shipping anything yet. Physical chip systems aren’t expected until 2028, with a real customer base targeted by 2030, so this is a bet on where the market goes, not a product announcement. Euclyd wants to sell hardware directly to companies that want their AI running on their own servers, and license the underlying design to other chipmakers too.

Nvidia isn’t going anywhere anytime soon, let’s be clear about that. But every dollar like this one is a sign that the biggest names in hardware think the AI chip market has room for more than one shape of winner. That kind of competition tends to be good news for anyone paying the electricity bill on all this compute.

What sticks with me is who’s writing the check. Samsung isn’t a random venture fund chasing a trend, it’s a company that actually knows how hard chips are to build at scale, and it’s choosing to back a totally different architecture instead of just riding the GPU wave. When the people closest to the manufacturing side start hedging their bets, that usually means they see the same crunch the rest of us are starting to feel, power, cost, and supply, all tightening at once.

——

Follow: @Ali Demi
Book your free AI clarity call, NOW!
https://buff.ly/TpWy277

——

Sources:
https://www.cnbc.com/2026/09/14/samsung-euclyd-ai-chip-funding.html
https://www.prnewswire.com/news-releases/euclyd-raises-over-200-million-to-break-the-ai-efficiency-wall-302878043.html
https://techfundingnews.com/samsung-backs-euclyds-e200m-round-to-challenge-nvidia-in-ai-chips/

Repost this. Thanks.