Samsung and Broadcom Sign $200 Billion AI Chip Partnership MOU Through 2030

MOU scope and financial scale

Samsung Electronics and Broadcom signed a memorandum of understanding to expand their strategic collaboration across memory and foundry technologies, supporting next-generation AI infrastructure, Samsung said in a statement reported over the weekend of July 25–26, 2026. The companies expect the collaboration to exceed $200 billion over the next five years through 2030, underscoring the scale of their joint efforts. Samsung described the expanded partnership as reflecting its focus on supporting customers with end-to-end semiconductor technologies across an increasingly diverse range of AI and high-performance computing applications.

Memory and HBM collaboration

On the memory side, Samsung and Broadcom plan to pursue a strategic partnership for the supply of industry-leading solutions, including High Bandwidth Memory, to power Broadcom's next-generation AI accelerators. The Seoul Economic Daily reported that memory supply contracts, previously on a one-year basis, are shifting to long-term agreements of five years or more. Samsung's memory chief Jun Young-hyun said last month he discussed next-generation foundry cooperation with Nvidia CEO Jensen Huang, including future HBM4E and HBM5 memory products, signaling parallel conversations across the custom-AI ecosystem.

Foundry and advanced packaging

In foundry, the collaboration focuses on Samsung's advanced 2-nanometer and below process technologies for Broadcom's products, including Wireless Broadband Communications solutions. The partnership is also expected to extend to advanced packaging technologies built on Samsung's 2nm process, such as 2.3D and 2.5D integration, enabling higher-performance and more energy-efficient AI and networking silicon. The deal provides for Broadcom's next-generation communications chips, designed for high-speed data transfer, to be made with Samsung's sub-2-nanometre process technology.

Strategic context and competitive positioning

Broadcom, one of the world's leading custom AI chip designers, brings expertise in designing application-specific integrated circuits for specific tasks, combined with Samsung's manufacturing capabilities. Winning long-term production commitments from Broadcom could boost utilisation at Samsung's advanced manufacturing facilities as the company seeks to narrow the gap with foundry leader TSMC. The tie-up comes as global technology companies increasingly develop their own custom AI accelerators rather than relying solely on general-purpose graphics processors, driving demand for specialised chip design and manufacturing partnerships. Samsung Electronics Chairman Lee Jae-yong met in rapid succession in Silicon Valley on July 25–26 with leaders from Nvidia, Broadcom, OpenAI and Anthropic to consolidate Samsung's memory position while exploring further foundry expansion, with the $200 billion Broadcom partnership viewed as a springboard for additional cooperation with other Big Tech companies.

Broader deal cluster and demand signals

The Broadcom MOU is part of a wider cluster of agreements. The Seoul Economic Daily reported that Samsung Electronics and SK hynix signed a series of AI chip supply contracts totaling about $950 billion (roughly 1,390 trillion won) with global big tech firms including Broadcom, Nvidia and Microsoft, with SK hynix's piece covering HBM4 supply to Nvidia and Microsoft. Separately, The Motley Fool reported on July 26 that Alphabet raised its 2026 data center capital expenditure guidance to $195 billion to $205 billion from a prior $185 billion to $195 billion range, with much of that spending flowing to a handful of suppliers including Broadcom, the design partner behind Alphabet's Tensor Processing Unit, and Nvidia. The simultaneous combination of long-term memory commitments and rising hyperscaler capex underscores sustained demand pressure on custom-AI silicon through the rest of 2026 and into 2030.