Samsung Just Backed a Nvidia Rival in a $230M round — Even Though Samsung Supplies Nvidia
A Dutch startup you've probably never heard of just raised €200M ($230M) — and one of the investors co-leading the round is a company that also makes memory chips for Nvidia itself.
Samsung co-led a Series A round for Euclyd, a two-year-old startup in the Netherlands building AI inference chips designed to rival Nvidia's GPUs. The round also drew Somerset Capital Partners, EQT's Scaleup Europe Fund, and Innovation Industries.
A few things worth sitting with:
→ Euclyd isn't just building a faster GPU clone — it's designing a different processor and memory architecture altogether, purpose-built for inference (the stage where a trained model actually generates answers for users, as opposed to training)
→ Samsung hasn't disclosed how much of the $230M is its own check — but its participation alone signals something
→ This lands right in the middle of a broader wave: multiple chip giants and tech majors are now funding GPU alternatives at the same time AI labs are publicly debating whether to slow the whole industry down
Here's what makes this interesting rather than just another funding round: Samsung is deeply embedded in Nvidia's own supply chain — it makes the memory that goes inside Nvidia's chips. Backing a company built to compete with Nvidia isn't a contradiction, it's a hedge. If inference workloads increasingly run on non-GPU architecture, Samsung wants exposure on both sides of that shift rather than being caught only serving the incumbent.
That's the quieter story running underneath every AI chip headline this year: the biggest players aren't betting on a single winner. They're placing chips (pun intended) across the entire board — training vs. inference, GPU vs. custom silicon, incumbent vs. challenger — because nobody actually knows which architecture wins the next five years.
Do you think purpose-built inference chips like this eventually chip away at Nvidia's dominance, or does Nvidia's software ecosystem keep it untouchable regardless of hardware alternatives?
#AI #Semiconductors #Nvidia #Samsung #TechFunding #AIChips #VentureCapital
— 𝔖𝔞𝔫𝔡𝔢𝔢𝔭 ℜ𝔞𝔦𝔷𝔞