Revenue forecast tops estimates as customers expand AI buildouts
Broadcom forecast second-quarter revenue above Wall Street expectations on Wednesday, pointing to sustained demand for advanced chips used in data centers that run artificial intelligence workloads. The outlook comes as large technology companies continue to commit substantial capital to AI infrastructure, supporting demand for semiconductors and related systems across the supply chain.
The company said it expects quarterly revenue of about $22.0 billion, above the $20.56 billion analyst consensus compiled by LSEG. Chief Executive Hock Tan said Broadcom’s AI-related momentum is strengthening, adding that AI semiconductor revenue is expected to reach $10.7 billion in the second quarter.
Big Tech firms including Alphabet, Microsoft, Amazon, and Meta are expected to spend at least $630 billion on AI infrastructure this year, a level of investment that can increase demand for chips, servers, storage, and networking gear supplied by Broadcom and peers.
New 10 billion buyback signals capital return alongside growth
Broadcom also announced a new share repurchase program of up to $10 billion, adding a shareholder return component at a time when investors across the AI supply chain are scrutinizing how companies balance reinvestment with capital returns. The company did not provide additional details in the information provided beyond the size of the authorization.
Shares were largely flat in volatile extended trading following the announcement. The stock is down about 8% so far this year after rising about 49% in 2025, a performance profile that reflects both the strength of the AI trade and periodic concerns about whether spending levels are sustainable.
Stacked chip design targets higher performance with lower energy use
Broadcom has been highlighting product development aimed at higher performance and better efficiency, a priority as data centers face growing power constraints. The company said last month it expects to sell at least 1 million chips by 2027 based on its stacked design technology, introducing a new product roadmap and a sales target that could translate into billions of dollars in revenue if reached.
Broadcom’s stacking approach is designed to help customers build chips with more computing capability while using less energy, aligning with the needs of AI software that requires rapidly increasing compute capacity.
Software segment slows as investors watch for AI bubble signals
Outside semiconductors, Broadcom’s infrastructure software business showed slower momentum. The segment grew about 1% to $6.80 billion in the first quarter, below analyst expectations for 2.6% growth to $6.88 billion, based on the figures cited.
Investor attention remains shaped by broader questions about the AI investment cycle. Last month, Nvidia reported better-than-expected results for the January quarter and issued above-consensus revenue guidance, yet the stock still sold off as markets debated whether AI spending is forming a bubble. Some investors have also expressed concern that Nvidia continues to deploy capital to expand the AI ecosystem with payoffs that are uncertain, rather than emphasizing shareholder returns.
Broadcom’s outlook suggests AI-related chip demand remains strong, but the market reaction indicates investors are weighing growth signals against valuation, capital intensity, and the durability of the current investment wave.