This is not investment advice. The author has no position in any of the stocks mentioned. Wccftech.com has a disclosure and ethics policy.
Microsoft backed OpenAI continues to remain interested in designing and developing its own chips in order to reduce the supply constraints that the broader industry is facing right now, according to a fresh report from The Information. OpenAI's narrow AI products require thousands of GPUs to function, and the details of the report share that the firm has talked with chip designer Broadcom which makes application specific chips for AI use cases. The portion of report shared by a Taiwanese publication outlines that while initially, OpenAI's CEO Sam Altman was interested in setting up his own chip manufacturing facility, his talks with Taiwanese contract semiconductor manufacturing giant TSMC made him change his mind.
Sam Altman Changes His Mind - Now Wants To Develop AI Chips Internally Says Report
According to the details, instead of considering setting up his own chip manufacturing facilities, Altman is now focused on creating a joint venture to develop the chips. His statements related to the funding required for the chip facilities had created quite a stir, as investors had balked at Altman's estimate of the $7 trillion figure that he would have needed to make all the chips required to power up AI software.
As part of efforts for a joint venture, Altman has reportedly held talks with Broadcom. Along with NVIDIA, Broadcom has been another hot AI chip stock due to its networking and ASIC products. These enable AI server connectivity and also allow developers to rely on customized chips suited to their needs. Broadcom's shares have gained 78% over the past twelve months, while NVIDIA's stock has gained 157% over the same time period.
While The Information's sources expect that OpenAI's custom AI chip will be similar in performance to NVIDIA's products, they do not expect it to be available until 2026 at the very least. Designing chips requires years of experience, and while the process has been somewhat streamlined due to the rise of contract chip manufacturers like TSMC, companies nevertheless have to devote significant time and efforts towards these endeavors.
Conversations with TSMC executives played a role in changing Altman's mind, and he has also recruited former Google employees that have worked on the search engine giant's custom AI chips. These talks have also focused on capacity allocation for OpenAI's products and away from NVIDIA's. TSMC, for its part, has shared willingness to allocate capacity if OpenAI commits to placing large orders reports UDN.
The discussions between Altman and Broadcom have also been reported by the Financial Times. However, its report stays mute on the TSMC angle. A sources quoted by FT shares that OpenAI "is not just going to sit back and let others build" the energy, chip and compute capacity required for AI workloads especially "when they are on the front line." Along with Broadcom, OpenAI has engaged with the broader industry according to the FT, with its source adding that these plans require "massive sums of money."
About the author: Ramish is a seasoned technology writer and editor with more than a decade of experience. He specializes in semiconductor fabrication and market analysis. With a background in finance and supply chain management - via his bachelors in Finance and a micromasters in supply chain management from MIT - Ramish combines financial rigor with deep industry insight to deliver accurate and authoritative coverage.
Follow Wccftech on Google to get more of our news coverage in your feeds.
