Broadcom Targets $230bn AI Chip Revenue by 2028

Broadcom shares rose approximately 2% on Monday as investors reassessed the company's custom silicon pipeline. The movement followed CEO Hock Tan's assurance that artificial intelligence infrastructure spending remains robust, countering recent market skepticism regarding the pace of AI development. Tan emphasized that expanding inference workloads will sustain demand for the firm's processors and networking products, directly supporting its core revenue streams.

The latest quarterly data underpins this confidence. Broadcom reported fiscal third-quarter revenue of $29.6 billion, marking an 86% increase year-over-year. Within this, AI semiconductor revenue surged 221% to $16.7 billion, highlighting the shifting mix of the company's business toward high-growth AI components rather than traditional networking hardware.

Anthropic becomes largest XPU customer

Management projects that Anthropic will emerge as Broadcom's largest XPU customer by 2027. This partnership is central to the company's long-term growth strategy, as it locks in significant future volumes of custom accelerators. By securing a leading AI developer as a primary client, Broadcom diversifies its dependency on a single hyperscaler and broadens its addressable market in the inference sector.

The firm outlines aggressive financial targets for its AI segment. AI semiconductor revenue is projected to reach approximately $115 billion in fiscal 2027. The figure doubles to an estimated $230 billion in fiscal 2028. These forecasts signal that Broadcom expects AI chip sales to constitute the majority of its total revenue within the next few years, fundamentally altering the company's financial profile.

Market reaction to AI spending

The recent share price rebound occurred after Broadcom and other AI-linked equities faced pressure. Market sentiment had wavered amid debates over the sustainability of future AI capital expenditure. Investors were concerned that slower model development cycles could reduce the need for new compute infrastructure, leading to a temporary de-rating of AI-heavy portfolios.

Tan's commentary aimed to mitigate these fears by linking demand to inference workloads rather than just training. As AI models are deployed at scale for real-time queries, the need for efficient, custom silicon increases. This shift suggests that Broadcom's product roadmap is aligned with the next phase of AI commercialization, providing a clearer line of sight to future revenue than previous quarters.

Strategic shift in revenue mix

The 221% growth in AI semiconductor sales illustrates a decisive shift in Broadcom's business model. The company is transitioning from a networking-centric provider to a dominant force in custom AI accelerators. This pivot is validated by the $16.7 billion quarterly figure, which now represents a substantial portion of total revenue, indicating that AI is no longer a secondary growth driver but the primary engine.

According to reports from tradingview.com, the market is increasingly focused on these long-term projections. The gap between current AI revenue and the $230 billion 2028 target implies a sustained period of high growth. For Broadcom, this means scaling manufacturing, design, and supply chain capabilities to meet the demand from Anthropic and other major AI labs, solidifying its position in the custom chip market.