Broadcom’s shares fell significantly in after-hours trading following a second-quarter earnings report that disappointed investors. The semiconductor manufacturer, known for its robust presence in the artificial intelligence (AI) market, posted revenue of $22.19 billion, which slightly trailed analysts’ consensus of $22.27 billion. The results come at a time when Broadcom is vying for a more substantial slice of the booming AI infrastructure sector. Currently, Nvidia holds a commanding lead in the market for graphics processing units essential for AI workloads.
Investor sentiment worsened after Broadcom provided a forecast for AI-related revenue that again undershot expectations. The company anticipates AI chip sales of about $16 billion for the third quarter, lower than the $16.36 billion sought by analysts from Visible Alpha. In a dramatic response, Broadcom’s shares plummeted more than 13% in extended trading, as market reactions reflected the company’s failure to upgrade its long-term growth outlook for AI revenue.
CEO Hock Tan indicated that Broadcom still expects to produce over 10 gigawatts of AI computing capacity by 2027—an uptick from prior estimates. However, the $100 billion revenue target for AI chips by that year remains unchanged. “Nothing slows down what was estimated prior – they just didn’t raise it,” remarked Ben Bajarin, head of Creative Strategies, underlining the gap between expectations and the reported metrics.
The heightened market sensitivity to AI semiconductor performance is significant, as many companies in the sector have experienced notable stock price increases over the last two years. As competition escalates in the custom AI chip market, firms like Marvell Technology are also catering to hyperscale cloud providers with tailored semiconductor solutions. Earlier this year, Marvell announced a forecast that its custom chip division might achieve over $10 billion in annual revenue by 2029, reinforcing the competitive landscape.
The demand for AI inference — the mechanism by which AI models generate user responses — has fueled a heightened need for customized processors optimized for specific tasks. In response, major cloud service providers are increasingly developing their own chips to enhance performance and lower costs, which has opened new opportunities for semiconductor manufacturers.
Concerns about Broadcom’s ability to meet future demand have been raised amid ongoing semiconductor supply chain issues. However, management during the earnings call assured investors that they are “very comfortable” with their supply capabilities for the years 2026 and 2027.
Market analysts noted that Broadcom’s slight revenue miss and substantial after-hours drop indicate that the market holds these semiconductor companies to exceedingly high standards. “Today’s miss on revenue and subsequent post-market pull back shows the market demands perfection for this chip rally to keep running,” stated Ryan Lee from Direxion.
Despite the overwhelmed response, Broadcom’s overall guidance for the upcoming quarter remains encouraging. The company estimates third-quarter revenue at about $29.4 billion, exceeding the consensus forecast of $28.54 billion from LSEG.
Underlying its earnings performance, Broadcom’s AI business remains a pivotal growth driver. The company’s solutions are in strong demand from hyperscale clients, such as Meta and Alphabet’s Google, which use Broadcom’s custom silicon to handle complex AI workloads. Industry projections suggest that the total expenditure of major technology corporations on AI infrastructure could surpass $700 billion this year, soaring from approximately $400 billion in 2025, marking a significant influx of investment into AI.
As the complexity of AI applications continues to advance, cloud providers are increasingly in need of specialized processors tailored for machine-learning tasks. Broadcom positions itself strongly in this arena, with plans to deliver around 10 gigawatts of AI computing capacity next year, anticipating even higher volumes in subsequent years.
For the second quarter, Broadcom reported AI-related semiconductor revenue of $10.8 billion—a year-over-year increase of 143%, surpassing forecasts driven by the demand for custom AI accelerators and networking solutions. This aligns with the company’s strategy to leverage its substantial capabilities in custom silicon design and advanced networking technologies in order to support the next wave of computing demands.
