Broadcom's CEO is sounding the alarm about a game-changer in the AI chip market

So far, Broadcom has been one of the biggest winners of the fourth-quarter 2024 earnings season (AVGO), whose market cap topped nearly $1 trillion last week.

Don’t miss the move: sign up for TheStreet’s FREE daily newsletter

Broadcom has had an outstanding year of consistent growth as a leader in the development and manufacturing of semiconductor technology. With gains of 118% over the year, driven primarily by advances in artificial intelligence (AI), AVGO stock is expected to end 2024 on an extremely strong note.

While Broadcom’s gains this year haven’t been as high as Nvidia’s (NVDA), the AI chip market may be facing a significant paradigm shift that could significantly benefit the former. On last week’s fourth-quarter earnings call, Broadcom CEO Hock Tan provided high revenue guidance for the next two years, highlighting an upcoming trend in the AI chip industry.

Broadcom CEO Hock Tan recently spoke about a new trend that could transform the AI chip market in 2025.NurPhoto/Getty Images

Over the past two years, industry-wide demand has pushed Nvidia to the forefront of the booming AI market. Many companies are heavily focused on developing large language models (LLMs), and Nvidia makes the only GPUs (graphics processing units) powerful enough to run many AI systems on them.

Related: Apple announces new partner to build AI chips, and it’s not Amazon

So far, no company has managed to produce a GPU that can really compete with Nvidia’s in terms of processing power and efficiency. However, on last week’s fourth-quarter earnings call, Tan discussed some elements that suggest the AI chip market may be changing in ways that benefit companies that produce custom silicon chips, particularly eXtreme Processing Units (XPUs ), unlike the general market purpose-built GPUs from Nvidia.

GPUs are considered extremely versatile and can typically handle a wide range of computing tasks. Nvidia’s chips are part of a robust software ecosystem, which has helped make them user-friendly.

In contrast, XPUs are highly specialized and often tailored to more specific, computationally demanding tasks, typically in the sense of AI and machine learning. They may be more limited in nature, but Broadcom’s fourth-quarter earnings report suggests specialized chips are gaining an edge over more general-purpose chips.

Related: Nvidia scales international hiring for key new tech sector

As noted, Tan sees significant revenue opportunities for Broadcom’s AI and AI networks, predicting a range of $60 billion to $90 billion in 2027, a significant increase from last year’s estimate of $15 billion to $20 billion.

The CEO attributed Broadcom’s strong AI growth to its custom XPUs and AI networks, pointing to a specific type of customer that seems to prefer these chips rather than the traditional GPUs.

“Certain hyperscalers have begun their respective journeys to develop their own custom AI accelerators, or XPUs,” he said, referring to cloud computing and data storage companies that do hyperscale computing. Broadcom’s hyperscaler partner list includes Amazon (AMZN) Web Services, Google (GOOGL) Cloud and Microsoft (MSFT) Azure blue.

Tan also announced that Broadcom has been selected to collaborate with two additional hyperscalers and is “in advanced development of their own next-generation AI XPUs.”

Broadcom’s success and growth plans make it clear that the market for custom silicon chips is growing. But if so, would companies that make commodity GPUs, particularly Nvidia, be threatened?

More tech stocks:

At first glance, it would be easy to assume that the increasing popularity of XPUs will threaten Nvidia’s market dominance. In July 2024 a report was published by Everything about industries examined this topic and discussed the utility of application-specific integrated circuits (ASICs), another custom silicon chip similar in functionality to the XPU.

Given the role XPUs played in Broadcom’s recent earnings slump, the question of custom silicon chips now seems even more relevant. However, one expert believes there is room in the market for both and that growing demand for both types of AI chips will fuel growth for both Broadcom and Nvidia in the coming year.

Ophir Gottlieb is CEO of Capital Market Laboratories, a sell-side research and financial technology company. He spoke with The street about Broadcom’s latest earnings release and what it’s likely to mean for chipmakers.

Related: Major Apple chip supplier expands into the US

“I think one type increases demand for the other, and it’s kind of a reinforcing cycle,” he says. “Not only is it not a zero-sum game, I think it’s synergistic.”

Gottlieb also explained the appeal of more specialized chips over Nvidia’s more general processors, emphasizing that leading technology companies run their AI platforms at their own software layer. In his words:

“If you have a custom software layer, you build the chips specifically to work with that software. And that’s why it’s more effective and efficient. It won’t be more efficient in the general sense.”

In a recent Broadcom note to investors, Gottlieb explains, “While NVIDIA dominates GPUs for AI, Broadcom’s custom silicon handles hyperscaler-specific workloads with unmatched efficiency,” reinforcing this bullish thesis about both companies.

Related: Veteran fund manager sees a world of pain coming for stocks

Related Posts

Golden State Warriors vs. LA Clippers December 27, 2024 Box Scores

What’s next with the labor dispute at Starbucks?

“Teen Konstas to open for Australia at Boxing Day India Test”

Leave a Reply Cancel reply