Islamabad, Sep 2: The state-of-the-art A16 chip from TSMC, which is being produced using its most sophisticated angstrom-level process, is creating a lot of talk even before it is officially put into mass production.
According to industry insiders, Apple, a significant customer of TSMC, has obtained the first batch of production capacity. It’s interesting to note that OpenAI, a prominent player in artificial intelligence, has also set aside A16 capacity. The long-term demand OpenAI has for specially created AI chips is what motivated this decision.
It is anticipated that the presence of a major player in the AI space will increase demand and visibility for TSMC’s AI-related business.
According to reports, OpenAI has changed how it plans to produce its unique AI chips. According to industry sources, the business first hired TSMC to set up a special wafer production plant. OpenAI, however, abandoned such intentions after evaluating the possible advantages.
Rather, the business has chosen a cooperative strategy, working with US firms such as Marvell and Broadcom to produce its ASIC chips. This is a major strategic move for Broadcom, since OpenAI is expected to rank among its top four clients.
TSMC will be significantly impacted by the partnership between OpenAI and American chip manufacturers Marvell and Broadcom. TSMC has been a longstanding customer of both Marvell and Broadcom. They are currently working with OpenAI to develop ASIC chips, and they plan to use TSMC’s advanced manufacturing methods to put them into production. According to the businesses’ chip design roadmaps, these chips are specifically anticipated to be produced utilizing TSMC’s 3nm family of processes and the forthcoming A16 technology.
TSMC is setting its sights on a major technological leap with its A16 process, the most advanced chip manufacturing technology the company has revealed to date. This process node marks TSMC’s entry into the “Angstrom era” of semiconductor fabrication, representing a new level of miniaturization.
Mass production using the A16 process is expected to commence in the second half of 2026, with TSMC’s Taiwan facilities leading the way.
