Euclyd raises €200M as Samsung funds and builds its rival chip for Nvidia inference

Euclyd, a chip company, has raised €200 million, and Samsung sits behind the round, funding the effort and building the rival inference chip aimed at Nvidia, according to a report from Tech Times.

The report pairs the two facts deliberately: the money and the silicon. With Samsung involved as both backer and builder, Euclyd has a path that goes beyond a paper design toward something that can actually be manufactured.

The target is inference, which has become the most contested stretch of the compute race. Training decides how capable a model can be; inference decides how cheaply and at what scale that capability can be served.

Nvidia's position in inference rests on its software ecosystem and on the momentum carried over from training. A challenger has to compete not on peak performance alone but on software maturity, migration cost for customers, and the ability to ship volume.

Samsung's motivation is straightforward. Backing a design house with a differentiated inference play gives it another card in its long-running contest with Nvidia, extending its reach from memory and foundry deeper into chip design.

The report does not give specifications for the chip, a production timeline, or named customers. €200 million is meaningful ammunition in the chip business, but it is a starting position rather than a final answer.

Two things are worth watching next: whether Euclyd publishes architecture details and benchmark numbers, and whether Samsung's role turns out to be capital, manufacturing, or both. The real test remains whether customers will move production inference workloads onto the part.