Broadcom guided for $56 billion in AI semiconductor revenue during fiscal 2026, representing approximately 180% growth from the prior year, CEO Hock Tan announced on June 3, 2026. The guidance underscores the company’s dominant position in custom silicon chips powering data center AI infrastructure as hyperscalers accelerate their spending on artificial intelligence.
The company’s second-quarter fiscal 2026 results delivered the strongest AI numbers in Broadcom’s history. AI semiconductor revenue hit $10.8 billion, up 143% year over year, on total revenue of $22.2 billion, up 48% annually. Adjusted earnings per share reached $2.44, topping analyst consensus of $2.40.
Broadcom is developing custom AI accelerators for six hyperscale customers, including Google, Meta, OpenAI, and Anthropic, alongside two additional unnamed partners. Each chip is engineered to one customer’s specific model architecture over an 18-to-24-month development cycle. The company disclosed that long-term supply agreements are in place extending into 2028, providing contracted demand visibility. Anthropic is scaling from roughly 1 gigawatt of compute in 2026 to more than 3 gigawatts in 2027, and OpenAI’s first-generation custom processor is slated for volume deployment in 2027 at over 1 gigawatt of capacity.
For the third quarter, Broadcom projects AI semiconductor revenue of approximately $16 billion, representing more than 200% year-over-year growth. That sequential step from $10.8 billion to $16 billion reflects confirmed deployments already in motion across the customer base. The company also reaffirmed its fiscal 2027 AI revenue guidance of in excess of $100 billion, signaling confidence in the multi-year demand cycle.
The guidance comes as the broader semiconductor industry experiences historic growth driven by AI. Global semiconductor revenues are forecast to reach $1 trillion or more in 2026, according to industry analysts. Broadcom holds approximately 60% projected market share in AI server compute application-specific integrated circuits (ASICs) by 2027, according to market analysis. The custom silicon model Broadcom operates creates structural advantages over sellers of off-the-shelf GPUs, since switching costs for customers are high once they commit to a proprietary architecture.
Despite the strong results, Broadcom stock declined in after-hours trading on June 3 because some investors had expected the company to raise its full-year AI guidance after first-half revenue approached $19 billion. The unchanged guidance was interpreted as a ceiling rather than a floor. However, the Q3 outlook of $29.4 billion in total revenue—up 84% year over year—came in above consensus estimates of approximately $28.5 billion, and free cash flow hit a record $10.26 billion in Q2, representing 46% of revenue.
Broadcom stock has surged on AI chip demand, gaining 24% year to date through early August 2026. The company’s custom chip strategy positions it to capture a growing share of hyperscaler capital expenditure. According to Wedbush analyst estimates, Big Tech’s capital spending is expected to touch $550–$600 billion in 2026, up from about $380 billion in 2025, with AI infrastructure investments driving much of the increase.
Sources
- Broadcom Inc. official press release — Q2 FY2026 financial results, AI revenue guidance of $56B for FY2026 (up 180%), Q3 outlook, and CEO Hock Tan statements
- Investing.com — Analysis of Broadcom’s AI revenue surge, custom chip customer details (Google, Meta, OpenAI, Anthropic), supply agreements through 2028, and market context
- DataForSEO search results — Industry forecasts for global semiconductor market reaching $1+ trillion in 2026, AI-driven growth, and Broadcom’s 60% market share projection in AI ASICs by 2027
