The chipmakers powering the artificial intelligence revolution have become stock market darlings. And while Nvidia is, thus far, the undisputed winner of the chip wars, it soon could be facing a robust challenger.
Broadcom, an American semiconductor manufacturer, announced a 10-to-1 stock split and better-than-expected earnings in its second-quarter earnings report on Wednesday. In the report, Broadcom also projected $51 billion in sales for the 2024 fiscal year, slightly higher than anticipated. Shares soared that day in after-hours trading and were going for about $1,740 around 3 p.m. ET on Friday—about 16% higher than their Wednesday opening price near $1,500.
In a note to investors on Thursday, Bank of America boosted its target price for Broadcom shares to $2,000 and raised its 2025 fiscal year sales forecast to $59.9 billion, which would be a 16% year-over-year gain. BofA cited semiconductor sales and its acquisition of software firm VMWare last year.
“We reiterate Buy, consider it a top AI pick” along with Nvidia, the analysts wrote, citing Broadcom’s potential growth in custom chips, ethernet networking, and VMWare upsells.
If BofA’s prediction materializes, that would push Broadcom’s market cap past $1 trillion and into an exclusive club currently occupied by just six firms: Microsoft, Apple, Nvidia, Amazon, Alphabet, and Meta.
Nvidia, valued near $3.25 trillion, is the only chip manufacturer to reach a 13-figure market cap. In just the past week, it overtook both Amazon and Alphabet, and last it week briefly nudged past Apple for the second spot behind Microsoft before retreating to third.
Nvidia volume today is about as much as the next 10 most traded stocks combined. I don't think I've ever seen that large of a gap bf, it's weird. pic.twitter.com/hJgprCM6vw
— Eric Balchunas (@EricBalchunas) June 5, 2024
The dominance of Nvidia, a company unknown to many consumers, is both simple and complicated. The short answer is that it produces chips that are becoming essential to every corner of the economy. The longer answer is that for the past 15 years or so, developers have built code around Nvidia’s CUDA architecture, making it difficult to unravel.
Last year, over $4 billion was invested into 93 separate efforts to do so, according to PitchBook. And earlier this year, the Unified Acceleration Foundation, or UXL, was formed by Intel, Google, Arm, Qualcomm, Samsung, and other tech companies to develop open-source software to free AI developers beholden to CUDA.