Competition for Nvidia? This chip company that hardly anyone knows is suddenly worth over $1 trillion.
Broadcom shares have skyrocketed recently, pushing the company’s market value to over $1 trillion.
Broadcom is critical for companies looking for alternatives to Nvidia’s dominance in AI chips.
Analysts say custom AI chips are on the rise and can strengthen technology companies’ negotiating positions.
The rise of artificial intelligence and the computing power it requires is putting many companies that have previously been under the radar in the spotlight. These days, Broadcom is pushing its way into the spotlight.
The stock has recently performed so well that the chip manufacturer entered the trillion club for the first time at the end of last week, reaching a market capitalization of over one trillion dollars (around 953 billion euros). The boost was triggered by an outstanding earnings report in which sales of customer-specific AI chips rose by 220 percent compared to the previous year.
Broadcom not only manufactures parts and components for data centers, but also develops and sells ASICs – application-specific integrated circuits – an industry abbreviation for customized chips.
Developers of custom AI chips, especially Broadcom and Marvell, are in a growth phase, according to Morgan Stanley.
Custom chips are gaining momentum
The biggest players in AI buy many chips from Nvidia, the $3 trillion giant with an estimated 90 percent market share in advanced AI chips.
But heavy reliance on a single vendor is not a comfortable situation for any company, which is why many large Nvidia customers also develop their own chips. But most technology companies don’t have large teams of silicon and hardware experts in-house. Of the companies they could turn to to develop a custom chip, Broadcom leads the way.
Although general-purpose chips such as Nvidia and AMD’s graphics processors are likely to retain the largest share of the AI chip market in the long term, custom chips are growing rapidly.
Morgan Stanley analysts predicted this week that the ASIC market will almost double to $22 billion (about €21 billion) next year.
According to analysts, much of this growth is due to Amazon Web Services’ Trainium AI chip . In addition, there are Google’s in-house AI chips, the so-called TPUs, in the production of which Broadcom is also involved.
In terms of the actual value of the chips deployed, Amazon and Google dominate. But OpenAI , Apple and TikTok parent company ByteDance are also reportedly developing chips with Broadcom.
ASICs bring negotiating power
According to the Morgan Stanley study, custom chips offer more value in terms of price-performance ratio.
ASICs can also be designed to perfectly fit the specific internal workloads of technology companies . The better these custom chips become, the more bargaining power they will have when negotiating with Nvidia to buy GPUs. But that will take time, the analysts write.
In addition to Broadcom, Silicon Valley neighbor Marvell is also seeing growth in the ASIC market, as are Asia-based companies Alchip Technologies and Mediatek, according to a statement to investors.
Analysts don’t expect custom chips to ever completely replace Nvidia GPUs, but without them, cloud service providers like AWS, Microsoft and Google would have much less negotiating power against Nvidia.
“In the long term, if cloud service providers do a good job, they could have greater bargaining power in sourcing AI semi-finished products with their own customized silicon,” Morgan Stanley analysts said.
Nvidia’s large R&D budget.
This is not necessarily bad news for Nvidia. The $22 billion market (about 21 billion euros) for ASICs is smaller than Nvidia’s revenue in just one quarter.
Nvidia’s R&D budget is huge, and many analysts are confident the company will be able to stay at the forefront of AI computing.
This is not necessarily bad news for Nvidia. The $22 billion market (about 21 billion euros) for ASICs is smaller than Nvidia’s revenue in just one quarter.
Nvidia’s R&D budget is huge, and many analysts are confident the company will be able to stay at the forefront of AI computing.
.
