Prime Highlights
- Hock Tan said the artificial intelligence boom could push Broadcom’s AI chip revenue beyond $100 billion by 2027, reflecting surging demand for custom AI hardware.
- The company is working with major technology firms to develop custom AI chips, supporting customers such as Google, Meta, Anthropic, and OpenAI.
Key Facts
- Broadcom reported first-quarter AI revenue of $8.4 billion, more than double the amount from the same period last year.
- The company posted total quarterly revenue of $19.3 billion, up 29% year-over-year, and expects AI semiconductor revenue to reach $10.2 billion in the current quarter.
Background
Hock Tan, the chief executive of Broadcom, expects the artificial intelligence boom to push the company’s AI chip revenue well beyond $100 billion by 2027. His comments came after the chipmaker reported strong financial results for the first quarter and issued an optimistic outlook for the coming months.
Speaking during the company’s earnings call, Tan said demand from large technology firms is increasing as more companies look to design their own custom AI chips. Broadcom supports these efforts by helping customers translate chip designs into working silicon before the chips are produced at major fabrication facilities such as Taiwan Semiconductor Manufacturing Company.
The company reported that AI-related revenue in the first quarter more than doubled from the same period last year to $8.4 billion. Total revenue rose 29% to $19.3 billion, beating market expectations. Broadcom also expects AI semiconductor revenue to reach $10.2 billion in the current quarter, showing continued strong demand.
Investors reacted positively to the outlook. Broadcom shares jumped more than 5% in extended trading after Tan discussed the company’s long-term growth potential driven by artificial intelligence.
Broadcom’s growth partly comes from its work with major technology companies building their own AI hardware. Tan said the company currently supports six key customers developing custom chips. These include Google, Meta, Anthropic, and OpenAI, with Fujitsu and ByteDance also expected to be part of the group.
Industry analysts say the company benefits as more technology giants shift toward in-house AI accelerators. For example, Google began developing its own tensor processing units in 2015 with support from Broadcom and later offered these chips to cloud customers, including Apple and Anthropic.
Despite the strong outlook, chipmakers still face several industry challenges. Companies must deal with shortages of high-bandwidth memory, which is essential for AI accelerators, along with manufacturing and packaging capacity limits at advanced semiconductor facilities.
Still, analysts note that Broadcom’s AI business goes beyond accelerators. The company also supplies networking switches, digital signal processors, and data processing units that support large-scale AI systems.
As AI adoption grows rapidly across industries, Broadcom expects demand for custom silicon and related infrastructure to continue expanding in the years ahead.
