AMD Holds Advanced Talks With Samsung For 2nm Production
According to Wccftech, citing a Korean outlet and unnamed industry sources, AMD is in "advanced talks" with Samsung Electronics' foundry to produce future AI-focused CPUs and accelerators on Samsung's 2nm process. The report says AMD CEO Lisa Su visited Samsung's Pyeongtaek fab in March, and the discussions reportedly cover orders for 2nm parts including codenames Venice and Verano, which Wccftech describes as a compute-optimized Zen 6 design and an agentic-AI variant respectively. Public coverage frames the outreach as part of broader capacity diversification as TSMC faces tight allocation for advanced nodes.
What happened
According to Wccftech, citing a Korean outlet (EDaily) and unnamed industry sources, AMD is in "advanced talks" with Samsung Electronics' foundry to manufacture next-generation AI CPUs and accelerators on Samsung's 2nm process. The report states that AMD CEO Lisa Su visited Samsung's Pyeongtaek foundry in March, and that discussions concern 2nm orders. The article names product codenames Venice and Verano, describing Venice as a compute-optimized Zen 6 offering and Verano as a variant targeted at agentic AI workloads; those product descriptions are reported by Wccftech.
Technical details
Per the Wccftech report, Samsung's foundry is offering a 2nm process that the article characterises as a GAA technology; that technical characterisation is attributed to the same reporting. The coverage frames Samsung as an alternative or backup source of advanced-node capacity for customers that have historically relied on TSMC.
Industry context
Editorial analysis: Companies in the server and datacenter silicon supply chain have increasingly explored multi-foundry sourcing to manage node-time-to-market and shortage risks. Observers following the sector have noted a broader pattern where customers place exploratory orders or enter early talks with multiple foundries when leading-edge capacity is constrained.
Significance for practitioners
Editorial analysis: For architects and procurement teams at cloud and hyperscaler customers, an additional qualified second-source at 2nm could alter validation windows, tapeout timing, and supply-risk assumptions even if production transitions typically require long qualification cycles and packaging revalidation.
What to watch
Editorial analysis: Watch for an official statement from AMD or Samsung confirming foundry agreements, published wafer/tapeout timelines, and any EDA or IP portability notes addressing Zen 6 variants. Also monitor TSMC capacity announcements and industry reporting on initial 2nm production ramp schedules, since those will determine whether multi-foundry sourcing is feasible for large-volume datacenter SKUs.
Scoring Rationale
This is a notable supply-chain development for datacenter silicon because credible multi-foundry discussions at 2nm could materially affect production timelines and risk management for AI CPUs. It is not yet confirmed, so practitioners should treat it as a developing capacity- and qualification-level story.
Practice interview problems based on real data
1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
