Broadcom CEO Hock Tan said the company’s AI Chip revenue will move significantly above $100 billion next year as demand for advanced silicon continues to surge. He made the prediction after Broadcom posted strong quarterly results and signaled ongoing momentum across the AI semiconductor market. The comment highlights how the AI Chip landscape is shifting as companies race to build more compute‑heavy systems.
Tan said demand for custom silicon is rising fast as major tech companies look for ways to support larger AI models and more intense workloads. He pointed to stronger engagement from several of Broadcom’s biggest customers, underscoring the accelerating need for reliable AI compute infrastructure. As a result of this trend, he expects the AI Chip sector to expand faster than earlier estimates suggested. Broadcom said its AI revenue more than doubled in the latest quarter, reaching $8.4 billion, while total sales climbed 29% to $19.3 billion. The company projected AI semiconductor revenue of $10.2 billion for the current quarter, reinforcing the strength of the AI Chip market. Shares rose more than 5% in extended trading after Tan’s remarks, reflecting investor confidence in the company’s AI roadmap.
The demand spike has not come without challenges. The industry continues to face shortages of high‑bandwidth memory and capacity constraints at advanced manufacturing nodes. These issues have slowed growth for several chip companies. However, Tan said Broadcom has secured the supply chain needed to meet its targets through 2027. He emphasized that the company worked closely with manufacturers to reduce bottlenecks and ensure stable production. Broadcom’s core strength now lies in custom AI accelerator design. The company helps customers translate architecture concepts into finished silicon before sending the designs to partners such as Taiwan Semiconductor Manufacturing Company for fabrication. This capability has become more valuable as more tech giants shift toward in‑house AI processors tailored to their workloads.
Tan confirmed that Broadcom is supporting six major customers as they design next‑generation chips for large‑scale AI compute. He highlighted Google, Meta, Anthropic and OpenAI as central partners driving custom silicon innovation, with Fujitsu and ByteDance expected to round out the group. Because custom AI deployment is now entering its next phase, Tan believes adoption will only accelerate from here.
Google was one of the earliest companies to build its own AI processors. Its Tensor Processing Units, developed with support from Broadcom, have been available to cloud customers since 2018. Broadcom said Google’s next‑generation chips are expected to generate even stronger demand in 2027, which aligns with the company’s long‑term AI Chip projections. Meta continues to develop its MTIA accelerator with Broadcom’s help. Some analysts questioned the future of Meta’s custom silicon program, but Tan insisted the MTIA roadmap is alive and well. He expects Meta’s growing focus on AI compute to fuel further expansion in the AI Chip market.
During the earnings call, analysts pressed Tan on the numbers behind the projected $100 billion milestone. He said the estimates reflect rising compute capacity across key customers, each requiring multiple gigawatts of infrastructure. While the dollars per gigawatt vary widely, Tan said the calculations behind his forecast are not far off. His comments show how scaling AI infrastructure drives both power demand and silicon demand in parallel.
Industry analysts noted that Broadcom’s AI Chip revenue includes more than AI accelerators. It covers digital signal processors, networking switches and data‑processing units that support AI workloads at scale. This broader product mix strengthens Broadcom’s position as a top supplier in the AI semiconductor ecosystem. In addition, the United States continues to play a central role in global AI expansion. U.S. hyperscalers remain the largest buyers of AI compute, and their rapid buildout of data centers is boosting demand for AI Chips across the country. This trend aligns with broader national goals to strengthen domestic semiconductor leadership. As the AI boom continues, Broadcom sees custom silicon as the industry’s next major growth engine. Tan said the company is preparing for sustained expansion as customers push deeper into AI‑heavy services. With the market shifting toward tailored accelerators and massive compute clusters, Broadcom expects its AI Chip business to remain a central force in the company’s long‑term growth.
