AI stocks may have lost some of their luster in 2026, but Broadcom ( AVGO ) management made a significant revelation during its fiscal 2026 first quarter earnings call that could change how investors view the company’s position in the AI boom.
Let’s find out what it is – and whether it’s still too late to get AVGO stock.
www.barchart.com
During the Q1 earnings call, CEO Hock Tan revealed that Broadcom is working with six customers with custom AI accelerators, calling the partnerships “deep, strategic and multi-year.” Tan also emphasized how focused the business is, pointing out that the company has “too few customers — six, to be exact” for the volume it’s pushing into AI. Broadcom has never officially disclosed the identity of its custom silicon customers. According to industry reports and analyst estimates, the company is collaborating with a number of prominent AI developers and hyperscalers — including Alphabet ( GOOGL ) Google, Meta Platforms ( META ), ByteDance, OpenAI, Anthropic, and Apple ( AAPL ) — on custom AI chips and data center architecture. These companies are among the world’s largest providers of AI infrastructure, and many are now building their own custom chips rather than relying on general-purpose GPUs. This is where Broadcom enters the picture. The company specializes in designing custom AI accelerators, often called XPUs, as well as high-performance network chips that connect thousands of processors inside large AI clusters.
Management emphasized that Broadcom is already deploying AI accelerators for many of its customers today, and that deployment is expected to grow dramatically over the next few years. This small group of large customers led to a 29% year-over-year (YOY) increase in revenue to $19.3 billion. Adjusted earnings rose 28% to $2.05 per share. The semiconductor segment led the way, generating revenue of $12.5 billion, up 52% YOY. Specifically, AI Semiconductor’s revenue rose 106% YOY to $8.4 billion. Broadcom expects the momentum to continue in Q2, generating $10.7 billion. AI network revenue also increased 60% YOY in Q1, accounting for a third of total AI revenue.
Broadcom’s Tomahawk switching chips and high-speed SerDes technology are gaining popularity among hyperscalers that build large AI clusters. This focused group of six AI customers is driving significant demand, giving Broadcom exceptional visibility into future revenue.
The highlight of the earnings call was Broadcom’s long-term projection for its AI chip business. The company now expects to generate more than $100 billion in AI chip revenue by 2027 alone. This estimate includes revenue from AI accelerators (XPUs), network switches, digital signal processors and other data center chips.
It’s worth noting that this $100 billion estimate is based on existing customers, namely the six largest customers, and current infrastructure plans. Future demand could be even higher. Management also emphasized that Broadcom has already secured key supply chain components such as advanced wafers, high-bandwidth memory, and substrates until 2028, ensuring it can meet the expected increase in demand.
During the earnings call, Broadcom management hinted at details that could be useful for investors who care and still believe in the growth of AI infrastructure. Broadcom said its $100 billion 2027 AI forecast is based only on chips, not racks or systems. This means that the total AI opportunity around Broadcom may be significantly larger than this number.
In addition, AI companies are starting to design chips with their own models, and Broadcom is a partner helping to build them. This will ultimately allow the company to quietly phase out traditional GPU-based infrastructure over time. Tan indicated that the development of AI infrastructure will not slow down and that it is only entering its next phase. Those who are patient and have a long investment horizon may want to watch their investment.
In fact, analyst estimates suggest that Broadcom’s revenue could reach around $104 billion in fiscal 2026 and possibly increase to $150 billion in fiscal 2027, while revenue is expected to grow rapidly over the next two fiscal years.
Despite all the benefits and significant growth that has propelled Broadcom to become a $1.57 trillion tech giant, its customer concentration carries some risks. Broadcom’s reliance on only a few customers introduces potential challenges that investors should consider.
If one or two key customers change their strategy or reduce their AI spending, Broadcom could suffer significantly. Major hyperscalers are known for their volatile capital spending cycles, and changes in investment priorities may have a significant impact on the semiconductor supply chain. In addition, many of these companies are developing their own AI chips, which may lead to more competition in the future.
Overall, the consensus for Broadcom stock is a “Strong Buy” rating. Of the 42 analysts with coverage, 36 rate the stock as a “strong buy,” three have a “moderate buy,” and three recommend a “hold” rating. AVGO stock is down 2.5% so far this year. Based on an average target price of $446.74, the stock has a potential upside of 35% from current levels. Additionally, the high price estimate of $525 indicates that the stock could gain as much as 59% over the next 12 months.
www.barchart.com
Broadcom’s deep partnerships with a small group of tech giants have put it at the center of the AI arms race, helping to design and supply the custom silicon that powers the next generation of AI systems. However, the same focus that creates opportunity also brings risk. Investors should now look to see whether Broadcom’s role as a custom chip architect for the AI industry will strengthen over time or whether an over-reliance on a few powerful customers will cause problems.
However, these six companies will play a major role in shaping Broadcom’s future.
As of the date of publication, Sushree Mohanty had no positions (either directly or indirectly) in any of the securities mentioned in this article. All information and data in this article is for informational purposes only. This article was originally published on Barchart.com