Broadcom CEO Hock Tan sees AI chip revenue ‘significantly’ above $100 billion next year

The rapid growth of artificial intelligence is creating massive opportunities for the semiconductor industry, and Broadcom is positioning itself at the center of this surge. CEO Hock Tan believes the company’s AI chip business could generate well over $100 billion in revenue in the coming years, driven by accelerating demand for custom silicon solutions.

Following stronger-than-expected financial results for the first fiscal quarter, Tan explained during the company’s earnings call that major technology clients are increasingly seeking Broadcom’s expertise to design specialized processors tailored for AI workloads. He also noted that the company has already secured the necessary supply chain capacity to support its ambitious revenue targets through 2027.

Broadcom’s performance highlights the momentum behind AI-driven computing. In the latest quarter, AI-related revenue more than doubled compared with the same period last year, reaching $8.4 billion, while overall company revenue climbed 29% to $19.3 billion. The company now expects AI semiconductor sales to reach around $10.2 billion in the current quarter.

Investors reacted positively to the outlook, pushing Broadcom’s stock up by more than 5% in after-hours trading after the earnings announcement.

Despite the strong demand outlook, the semiconductor industry still faces several operational challenges. Companies such as Broadcom have been dealing with limited supplies of high-bandwidth memory, a key component for AI accelerators, along with capacity constraints in advanced chip manufacturing and packaging processes.

Broadcom plays a crucial role in helping technology companies convert their processor designs into physical chips. The company provides critical engineering and backend design support before the chips are manufactured in fabrication facilities run by firms such as Taiwan Semiconductor Manufacturing Company.

This specialized service has fueled Broadcom’s growth as large technology companies increasingly develop their own AI accelerators. Tan noted that the deployment of custom AI chips is entering a new phase of expansion, with Broadcom currently assisting six major clients in designing advanced processors. Among the key partners are Google, Meta Platforms, Anthropic, and OpenAI, with Fujitsu and ByteDance expected to join the group.

Google has been a pioneer in designing its own AI processors, first introducing its Tensor Processing Units in 2015 with Broadcom’s assistance. Since 2018, Google has also made these chips available to cloud clients, with organizations such as Apple and Anthropic among the users. Broadcom anticipates even stronger demand once Google’s next-generation chips arrive around 2027.

Meanwhile, Meta continues to develop its own MTIA AI accelerator with Broadcom’s support, even as some analysts question the long-term prospects of the project. Tan emphasized that Meta’s chip roadmap remains active and progressing.

During the earnings call’s Q&A session, Bernstein Research analyst Stacy Rasgon asked Tan about the potential sources behind the projected $100 billion in AI-related revenue. Rasgon estimated significant computing capacity commitments from companies such as Anthropic, Google, Meta, and OpenAI. Tan responded that revenue levels per gigawatt of capacity can vary widely but indicated that the analyst’s estimates were generally close.

Although AI accelerators are driving much of the growth, Broadcom’s AI-related business extends beyond just those chips. According to industry analyst Ben Bajarin of Creative Strategies, the company’s AI portfolio also includes components such as digital signal processors, data processing units, and networking switches — all critical elements supporting modern AI infrastructure.

Voice Of Osiz

The growing demand for AI chips highlighted by Broadcom reflects the rapid expansion of intelligent computing across industries. At Osiz, we see this as a strong indicator that AI infrastructure is becoming the backbone of next-generation digital ecosystems. Custom silicon development for AI workloads is enabling faster processing, improved scalability, and highly optimized performance. As major technology leaders like Google and Meta Platforms continue investing in specialized processors, the demand for advanced AI-driven solutions will only accelerate. This momentum signals a broader transformation in how enterprises build and deploy AI at scale. Strategic chip innovation will play a crucial role in powering future AI platforms. At Osiz, we believe such advancements will unlock new possibilities for intelligent applications across global industries.

Source: CNBC