Altman Meets Asian Chip Suppliers

OpenAI CEO Sam Altman visited South Korea and Taiwan as the company advances development of its own AI chips, meeting semiconductor executives and suppliers to discuss cooperation. In South Korea, he met Samsung Electronics executives for the first time in about a year. Richard Ho, OpenAI’s vice president overseeing in-house AI chip development, accompanied him, prompting speculation about Samsung’s possible role in next-generation AI accelerators.

Seoul Economic Daily reports, citing industry sources, that OpenAI’s Jalapeño accelerator is expected to use Samsung’s next-generation HBM4 memory. Samsung signed a memorandum of understanding with Broadcom in July for up to $200 billion in memory and foundry cooperation through 2030. Because Broadcom is developing custom AI chips for OpenAI, Samsung could also have opportunities in foundry and packaging services. Altman also met SK Group Chairman Chey Tae-won, with SK hynix and SK Telecom executives reportedly present. Business Korea says the undisclosed discussions are believed to have covered existing HBM supply cooperation and AI data center development in South Korea, as well as possible future collaboration.

OpenAI’s Stargate team reportedly visited South Korea before Altman to discuss semiconductor supply volumes, contract terms and AI infrastructure development with local companies, including HBM supply talks with Samsung, according to The Asia Business Daily. The meetings came amid constrained HBM supply and rising costs. TrendForce projects HBM’s blended average selling price will rise 121% year over year in 2027, citing supply constraints and a growing share of higher-priced HBM4.

In Taiwan, Foxconn Chairman Young Liu confirmed meeting Altman, and Reuters reports Foxconn is working on a new collaboration with OpenAI. Commercial Times also reports that Altman met TSMC Chairman C.C. Wei and engaged with AI server suppliers Quanta and Wistron. OpenAI and Broadcom unveiled Jalapeño, their first custom AI inference chip, in June 2026; it took nine months from initial design to tape-out. OpenAI says the chip has undergone performance testing, with initial deployment planned by the end of 2026 and expansion to gigawatt-scale capacity in the coming years. Commercial Times, citing supply chain sources, reports that the first-generation ASIC uses TSMC’s 3nm process, while the second-generation chip is reportedly planned for A16, which uses backside power delivery to improve performance and energy efficiency.

OpenAI CEO Sam Altman visited South Korea and Taiwan as the company advances development of its own AI chips, meeting semiconductor executives and suppliers to discuss cooperation. In South Korea, he met Samsung Electronics executives for the first time in about a year. Richard Ho, OpenAI’s vice president overseeing in-house AI chip development, accompanied him, prompting speculation about Samsung’s possible role in next-generation AI accelerators.

Seoul Economic Daily reports, citing industry sources, that OpenAI’s Jalapeño accelerator is expected to use Samsung’s next-generation HBM4 memory. Samsung signed a memorandum of understanding with Broadcom in July for up to $200 billion in memory and foundry cooperation through 2030. Because Broadcom is developing custom AI chips for OpenAI, Samsung could also have opportunities in foundry and packaging services. Altman also met SK Group Chairman Chey Tae-won, with SK hynix and SK Telecom executives reportedly present. Business Korea says the undisclosed discussions are believed to have covered existing HBM supply cooperation and AI data center development in South Korea, as well as possible future collaboration.

OpenAI’s Stargate team reportedly visited South Korea before Altman to discuss semiconductor supply volumes, contract terms and AI infrastructure development with local companies, including HBM supply talks with Samsung, according to The Asia Business Daily. The meetings came amid constrained HBM supply and rising costs. TrendForce projects HBM’s blended average selling price will rise 121% year over year in 2027, citing supply constraints and a growing share of higher-priced HBM4.

In Taiwan, Foxconn Chairman Young Liu confirmed meeting Altman, and Reuters reports Foxconn is working on a new collaboration with OpenAI. Commercial Times also reports that Altman met TSMC Chairman C.C. Wei and engaged with AI server suppliers Quanta and Wistron. OpenAI and Broadcom unveiled Jalapeño, their first custom AI inference chip, in June 2026; it took nine months from initial design to tape-out. OpenAI says the chip has undergone performance testing, with initial deployment planned by the end of 2026 and expansion to gigawatt-scale capacity in the coming years. Commercial Times, citing supply chain sources, reports that the first-generation ASIC uses TSMC’s 3nm process, while the second-generation chip is reportedly planned for A16, which uses backside power delivery to improve performance and energy efficiency.