On Feb. 6 (local time), SK Group Chairman Chey Tae-won met with Broadcom CEO Hock Tan at Broadcom headquarters in San Jose, California. During the meeting, the two leaders discussed mid-to-long-term memory market forecasts, supply strategies, and the companies’ respective investment portfolios. The visit was part of broader efforts to further expand the AI memory technology leadership spearheaded by SK hynix in collaboration with global partners.
Amid structural uncertainty and volatility in memory supply and demand driven by the recent surge in global AI demand, the two companies reviewed their existing cooperation framework centered on HBM for data centers and shared measures to stabilize the supply of semiconductors for the broader AI infrastructure ecosystem. SK hynix emphasized that its accumulated experience in mass-producing HBM and quality competitiveness ensure the stable supply of memory to Broadcom’s key global customer projects.
Notably, the meeting focused on next-generation AI chip architectures and memory integration strategies, building on the two companies’ established HBM technology collaboration. With Broadcom providing customized AI accelerators and networking solutions to global Big Tech customers and SK hynix maximizing system performance through HBM, the two parties engaged in in-depth discussions on new design and packaging approaches for the simultaneous optimization of next-generation HBM and AI-specific chips.
During the meeting, SK hynix shared its future HBM roadmap and technology development direction, agreeing to expand the scope of cooperation so that the company’s memory technology can be integrated at the early design stages of Broadcom’s AI chips. Both parties committed to maintaining a “one-team” approach in the global market, positioning themselves as key pillars of next-generation AI infrastructure.
