Cristiano Amon, Qualcomm CEO, has kicked off a high-stakes diplomatic push in Seoul by holding back-to-back meetings with Samsung Electronics, SK hynix, and LG Electronics, signaling a concerted effort to lock down Korea’s AI-chip ecosystem from foundry to physical AI hardware. The visit underscores Qualcomm’s pivot from a smartphone-centric AP leader to a full-stack AI player spanning on-device intelligence, automotive, and data-center inference.
2nm AP and Samsung Foundry – Snapdragon 8 Elite 2 at Stake
Amon’s first-stop discussions centered on Samsung’s cutting-edge 2-nanometer process and the potential foundry production of Qualcomm’s next-generation application processor, the Snapdragon 8 Elite 2.
Industry sources indicate that the talks included concrete scenarios for outsourcing design-complete Snapdragon 8 Elite 2 tape-outs to Samsung’s leading-edge logic node, dovetailing with Amon’s earlier CES 2026 remarks that advanced-node contract-manufacturing talks with Samsung were already underway.
For Samsung Foundry, this represents a strategic win securing a flagship mobile-AP program at 2nm strengthens its competitive posture against TSMC in the AI-enabled smartphone segment, where high-performance CPU-GPU-NPU triads are now central to on-device AI workloads.
For Qualcomm, the collaboration reduces supply-chain risk and ensures tighter co?design between modem-radio stacks and Samsung’s FinFET-based compute blocks.
SK hynix and Memory: Securing AI-Grade DRAM and HBM
Alongside foundry talks, Cristiano Amon met SK hynix executives to address memory-supply strategies in the face of Qualcomm’s expanding AI footprint. Discussions reportedly spanned server-grade DRAM and advanced AI memories such as high-bandwidth memory (HBM) and associated compression-attached memory modules, all of which are critical for Qualcomm’s data-center AI inference play and on-vehicle AI workloads.
SK hynix’s leadership in HBM and its recent ramp-up of SOCAMM-class memory for Nvidia’s next-gen AI chips makes it a natural anchor partner for Qualcomm’s AI200-style accelerator roadmap. By aligning with SK hynix, Qualcomm aims to secure a stable, high-bandwidth memory pipeline that can scale across smartphones, automotive domain controllers, and AI?inference server nodes.
LG Electronics and Physical AI: Cars, Homes, and Robotics
Later that day, Qualcomm CEO held a closed-door meeting with LG Electronics CEO Ryu Jae-cheol and other senior executives, focusing on “physical AI” and automotive electronics.
Physical AI refers to AI-driven systems embedded in real-world environments vehicle cockpits, smart home hubs, robotics platforms, and connected appliances where Qualcomm’s wireless, audio, and compute stacks converge with LG’s hardware and software ecosystems.
LG is aggressively expanding beyond white-goods into mobility, telematics, and robotics, making it a logical partner for Qualcomm’s Dragonwing IQ10-class robotics processors and 5G-connected cockpit platforms.
The Korea-based handset-to-vehicle-to-home continuum also plays into Qualcomm’s broader “AI everywhere” strategy, where multiple edge nodes share a common AI?core architecture and software stack.
Why This Seoul Shuttle Matters for the AI Ecosystem
Taken together, Amon’s Seoul itinerary paints a picture of Qualcomm building a “vertical AI alliance” across Korea’s semiconductor and electronics stack, Samsung Foundry for 2nm APs, SK hynix for memory, and LG Electronics for end-device AI.
This tripartite structure closely mirrors the broader industry trend where AI-chip design houses are no longer just tape-out partners but full-stack ecosystem orchestrators, tying together nodes, memory, and physical AI applications.
