Broadcom reassured investors about artificial intelligence (AI) chip demand on Thursday, issuing a robust second-quarter forecast and hinting at new hyperscale customers that could further bolster revenue in a competitive semiconductor market.
The chipmaker projected revenue of approximately $14.90 billion for the quarter, surpassing LSEG estimates of $14.76 billion. The announcement alleviated fears about slowing AI chip demand, which had been sparked by Marvell Technology’s (MRVL.O) weaker outlook a day earlier. Broadcom shares, which had closed down 6%, surged 14% in extended trading following the earnings report.
CEO Hock Tan stated that revenue from AI semiconductors is expected to reach $4.4 billion in the second quarter, driven by cloud computing companies investing in custom AI chips as they expand their data centers. Broadcom currently supplies three major hyperscale customers but has engaged with four additional firms seeking to develop their own AI chips. These new customers are not yet included in Broadcom’s projected AI revenue opportunity of $60 billion to $90 billion by 2027.
Last month, Reuters reported that OpenAI was working with Broadcom to finalize its first custom chip design, aiming to reduce dependence on Nvidia (NVDA.O). Analysts see Broadcom benefiting as large tech firms shift from off-the-shelf chips to in-house processors tailored for complex AI workloads.
Broadcom also reported first-quarter revenue of $14.92 billion, exceeding analyst expectations of $14.61 billion. AI-related revenue surged over 77% to $4.1 billion, while its infrastructure software segment grew more than 47% to $6.70 billion, surpassing forecasts. The company is further evaluating Intel’s 18A manufacturing process for future chip production.