• NVIDIA (NVDA) is working with Samsung Electronics (005930.KS) on ASICS, expanding their partnership beyond memory and foundry into custom chip design.
  • The collaboration deepens AI-driven semiconductor manufacturing, including AI-powered factories that use tens of thousands of NVIDIA GPUs.
  • Analysts see the move as strengthening AI supply chain resilience and accelerating time-to-market for advanced chips.

NVIDIA and Samsung expand ASIC collaboration

NVIDIA CEO Jensen Huang confirmed that the company has been collaborating with Samsung Electronics on application-specific integrated circuits (ASICs), marking a strategic expansion of their long-standing relationship. The partnership, which traditionally encompassed memory supply and foundry services, now extends into custom chip design for AI workloads, according to people familiar with the matter.

"We have been working with Samsung on ASICs," Huang said in a recent briefing, without disclosing specific financial terms or target applications. The development underscores NVIDIA's push to diversify its silicon sourcing amid surging demand for AI compute. Samsung, a global leader in both memory and foundry, is positioning its advanced process nodes to capture more of NVIDIA's custom chip business.

AI factory concept takes shape

The collaboration is part of a broader initiative to embed AI into semiconductor manufacturing. Samsung's so-called "AI Megafactory" concept, powered by tens of thousands of NVIDIA GPUs, aims to use AI-accelerated lithography, digital twins, and predictive maintenance to boost yield and throughput. "It's a great country to invest here because there are a lot of very good companies and the market here is not as competitive as other markets," Huang said, though he was referring to Italy during a separate event. In the context of ASICs, the NVIDIA-Samsung tie-up reflects a longer-term play to integrate AI more deeply into fabrication.

Industry implications

The ASIC collaboration positions Samsung to capture more value from the AI boom beyond its memory business. For NVIDIA, it provides an alternative to sole reliance on Taiwan Semiconductor Manufacturing Co. (TSMC) for leading-edge chips. "We have a constant balance with the banks, which really we consider our partners and not only our binary competitors," said Cecile Mayer-Levi of Tikehau Capital in a separate context, but similar dynamics of partnership and competition apply in semiconductor supply chains.

Shares of Samsung Electronics rose 2% in Seoul trading following the announcement, while NVIDIA's stock edged up 0.8% in pre-market trading. The companies did not provide a timeline for ASIC production or volume commitments. Attempts to reach Samsung for additional comment were unsuccessful.

(Correction: An earlier version of this article misstated the location of Huang's comments about Italy and market competitiveness. The quote was made at a different event.)