OpenAI is preparing to launch its inaugural artificial intelligence chip next year through a collaboration with U.S. semiconductor leader Broadcom, according to the Financial Times on Thursday, citing sources familiar with the matter.
The tech firm intends to deploy this chip for its internal operations, rather than offering it to outside clients, the FT reported, referencing an individual close to the project. OpenAI and Broadcom did not immediately respond to Reuters’ inquiries following standard business hours.
OpenAI, widely recognized for popularizing generative AI capable of producing human-like answers to questions, depends on immense computing resources to train and operate its systems.
Last year, Reuters revealed that OpenAI was partnering with Broadcom and Taiwan Semiconductor Manufacturing Co to develop its first proprietary chip to support its AI systems, while also continuing to use AMD and Nvidia chips to meet rising infrastructure requirements.
At that stage, OpenAI had been exploring several alternatives to diversify its chip sources and curb expenses. In February, Reuters reported that OpenAI was advancing its initiative to reduce dependency on Nvidia by developing its initial generation of in-house AI processors.
Sources told Reuters that the company was finalizing the design of its first proprietary chip and intended to send it to TSMC for production in the upcoming months.
Broadcom Sees Strong AI Growth Ahead
Broadcom’s CEO, Hock Tan, stated on Thursday that the company anticipates artificial intelligence revenue in fiscal 2026 to “improve significantly,” following more than $10 billion in AI infrastructure orders from a new customer, though the firm did not disclose the name.
Tan explained during an earnings call that a new prospect became a qualified customer by placing a firm order last quarter.
Earlier this year, Tan had mentioned four potential clients who were “deeply engaged” in developing their own custom chips, supplementing Broadcom’s three existing large clients.
OpenAI’s push to create a proprietary AI chip mirrors strategies adopted by other tech giants such as Google, Amazon, and Meta, all of which have designed custom silicon to manage AI workloads.
This trend comes as the demand for computing power necessary to train and operate AI models continues to rise sharply, underscoring the growing emphasis on tailored infrastructure within the industry.
Ready to take your career to the next level and stay competitive in your field?
Check out our handpicked collection of certification prep resources, designed to help you strengthen your skills, optimize your workflow, and stay ahead of the curve:
- CompTIA Exam Prep – Develop expertise in IT support, networking, and cybersecurity.
- HR Certification Resources – Deepen your knowledge in essential HR areas like compliance, workforce planning, and talent management.
- PMP Study Materials – Boost your project management capabilities, covering budgeting, scheduling, and team leadership.
- Praxis Test Prep – Get fully prepared for your teaching certification with thorough guides, practice tests, and materials aligned with state standards.