2 Sources
2 Sources
[1]
OpenAI's Sam Altman had secret TSMC meeting over future chip supply, report claims -- AI pioneer in Asia as South Korea confirms 20MW data center deal with ChatGPT maker
He followed it up with a trip to South Korea to announce data center builds with Samsung and SK Hynix. OpenAI CEO Sam Altman reportedly took a quiet trip to Taiwan this week to hold meetings with representatives of TSMC and Foxconn and discuss collaborating on chip design and manufacturing, as well as AI server infrastructure development, according to Digitimes. These kinds of partnerships will be crucial if OpenAI is to meet its commitments as part of major initiatives like Stargate, to build hundreds of billions of dollars worth of data centers and "AI factories," over the next few years. Although not announced by any of the involved companies or Altman himself, there's no denying he's in the region, as it was confirmed just one day later, Wednesday, September 30, that Altman had been in South Korea meeting with President Lee Jae Myung. Korea Times reported that he secured non-binding deals to build a 20-megawatt data center in Phang in partnership with Samsung, and another in the South Jeolla province with SK Hynix. The alleged meetings in Taiwan were arguably more important, however. Although Altman and OpenAI have been doing deals around the world to build out AI infrastructure, it needs many hundreds of thousands of chips to do it with, and TSMC and Foxconn are set to be major suppliers. TSMC is the world's largest and most advanced chip design and manufacturing company, while Foxconn is the largest supplier for Oracle, the cloud computing company that OpenAI has struck a deal worth $300 billion for compute power. Foxconn is an integral partner with Japanese investment vehicle Softbank, which is also heavily invested in OpenAI and a range of its data center projects. Foxconn will manufacture hardware for Softbank at the Ohio facility it sold to the company earlier this year. Altman's meetings with TSMC and Foxconn, if real and fruitful, likely further entrenched their cooperation on scaling up OpenAI's global AI infrastructure, as well as gaining deeper insight into TSMC's planned advanced process nodes for future chip designs. Another area likely discussed, and potentially more impactful for the future, is OpenAI's own chip design ambitions. As it stands, it relies on hundreds of thousands of expensive and power-hungry Nvidia GPUs. Alongside China and other countries and companies, OpenAI wants to untether itself from being so reliant on Nvidia moving forward, so it is developing its own AI application-specific integrated circuit (ASIC) chips for AI inference. OpenAI reportedly formed an ASIC design team in 2024 and is alleged to be working with Broadcom to develop a custom AI chip on TSMC's advanced 3nm process. Combining high bandwidth memory and advanced packaging technologies, OpenAI's bespoke hardware is expected to enter mass production in Q3 2026, following a recent delay.
[2]
OpenAI's CEO Sam Altman Pays a 'Secret Visit' To Taiwan, Reaching Out to TSMC For Its In-House AI Chip Which is Slated For 2026
OpenAI's CEO has recently visited Taiwan, making crucial stops at TSMC and Foxconn to discuss the company's AI projects, including Stargate and its pursuit of ASICs. OpenAI is currently in the pursuit of rapidly expanding its compute capabilities, whether it is striking 'multi-billion' dollar deals with the likes of NVIDIA and Coreweave for AI hardware, or even working towards its custom AI chips. Now, in a new report by the Taiwan Economic Daily, it is revealed that OpenAI's Sam Altman has paid a visit to Taiwan, discussing matters related to its highly ambitious Stargate project with Foxconn, as well as visiting TSMC to explore their self-built AI chip plans. For those unaware, Foxconn is set to be one of the largest suppliers for NVIDIA's rack-scale solutions, which will power the Stargate project. This project plans to open several data centers, involving partners such as OpenAI, Oracle, and SoftBank. Stargate is valued at around $500 billion and is projected to be America's biggest venture in terms of computing power. Altman's visit to Foxconn would likely be to discuss AI server production capacities, as well as how the Taiwanese firm would contribute to developing the mega AI infrastructure. The more important visit is around Altman and TSMC, as the report indicates that OpenAI's CEO has discussed the company's pursuit of developing AI chips. Initially, the custom AI chip project was expected to be co-developed with Broadcom. However, now that AI giant OpenAI has recruited individuals from Google's TPU project, it appears that OpenAI will likely take chip design into its own hands, eventually putting it up for manufacturing at TSMC. Specifications of OpenAI's custom chip are uncertain; however, it is speculated that the firm will utilize TSMC's 3nm process, with integration expected by 2026. OpenAI's ASIC will mark the first 'demonstration' of custom chip projects from Big Tech that could pan out, relative to NVIDIA's offerings, hence it is undoubtedly a huge deal.
Share
Share
Copy Link
OpenAI CEO Sam Altman's recent trip to Asia involves crucial meetings with TSMC and Foxconn in Taiwan, followed by data center deals in South Korea. These moves are part of OpenAI's ambitious plans for AI infrastructure expansion and custom chip development.
OpenAI CEO Sam Altman has been making significant moves in Asia, with a series of high-profile meetings and deals that could shape the future of AI infrastructure and chip development. The AI pioneer's recent activities include a reported secret visit to Taiwan and confirmed meetings in South Korea, all aimed at bolstering OpenAI's technological capabilities and global presence .
According to reports, Altman made an unannounced trip to Taiwan, where he held crucial meetings with representatives from Taiwan Semiconductor Manufacturing Company (TSMC) and Foxconn . These discussions reportedly centered around potential collaborations in chip design, manufacturing, and AI server infrastructure development . The meetings are seen as critical for OpenAI's ambitious plans, including the Stargate initiative, which aims to construct hundreds of billions of dollars worth of data centers and "AI factories" in the coming years .
Following his Taiwan visit, Altman's presence in the region was confirmed when he met with South Korean President Lee Jae Myung . During this visit, OpenAI secured non-binding agreements to construct two 20-megawatt data centers in South Korea . One center is planned for Phang in partnership with Samsung, while another is set to be built in the South Jeolla province in collaboration with SK Hynix .
A key focus of Altman's meetings, particularly with TSMC, appears to be OpenAI's ambitions to develop its own AI chips . The company reportedly formed an ASIC (Application-Specific Integrated Circuit) design team in 2024 and is working towards creating custom AI chips for inference tasks . These chips are expected to be manufactured using TSMC's advanced 3nm process, with mass production potentially beginning in Q3 2026 .
Related Stories
OpenAI's push for custom chip development is partly driven by a desire to reduce its reliance on Nvidia GPUs, which are currently essential for its operations . By developing its own ASICs, OpenAI aims to create more efficient and tailored solutions for its AI infrastructure needs .
These strategic moves by OpenAI underscore the company's commitment to expanding its global AI infrastructure and technological capabilities. The partnerships and deals being forged in Asia are likely to play a crucial role in OpenAI's future developments, potentially reshaping the landscape of AI computing and infrastructure on a global scale .
Summarized by
Navi
10 Feb 2025•Technology
28 Sept 2024
30 Oct 2024•Technology