Nvidia Isn’t the Only Way to Play AI. The List of Possible Winners Is Long.

But there are many other theories on who the winners will be. On Monday, three Wall Street research firms weighed in on possible winners—and losers—from AI.

The most comprehensive of these new report, at 184 pages, comes from TD Cowen analyst John Blackledge. He writes that the generative-AI trend will be “a monumental shift,” on par with the adoption of the internet. Cowen is modeling $81 billion in generative AI software spend in 2027, a 190% 5-year compounded growth rate. 

Cowen’s top three AI plays are obvious ones:
Alphabet
(GOOGL),
Microsoft
(MSFT), and
Nvidia
(NVDA). Alphabet and Microsoft have a clear and early advantage in the AI software market—and both have begun adopting AI features across their software portfolios, going far beyond search. And Nvidia is the clear leader in chips used to train large language models. All have surged higher this year.

Blackledge adds other names to the list, though.

Amazon
(AMZN), which is working on large language models of its own, provides a comprehensive set of AI tools to Amazon Web Services customers, and should benefit from the increase in demand for cloud-computing services that the expanded adoption of AI should trigger. And
Meta
(META) should be a beneficiary, as Facebook’s parent ramps up spending on generative-AI software. In particular, the analyst thinks AI should boost the efficiency of Meta’s advertising sales across the company’s social networks.

Blackledge also expects a boost for
Adobe
(ADBE), which is already showing off new generative-AI content creation tools.
Oracle’s
(ORCL) cloud-computing platform, he wrote, “has some competitive advantages over the incumbents in capturing this next wave of cloud workload growth” driven by AI and machine learning. Other picks include the chip makers Advanced Micro Devices (AMD) and Micron, the data center operators
Digital Realty Trust
(DLR) and
Equinix
(EQX), and security software stocks
CrowdStrike
(CRWD) and
SentinelOne
(S).

Needham analyst N. Quinn Bolton offered up a diverse list of potential winners.

Echoing a theme that Goldman’s Dane pointed out last week on Barron’s Live, the Needham note positions the chip industry as “the picks and shovels” of AI computing. The trend should drive demand not only for Nvidia’s GPUs, but also for sensors, memory, networking chips, and various other parts.

Needham’s list of winners includes not only Nvidia, which he considers “the best positioned company” in AI, but also Micron,
Cerence
(CRNC), which makes voice-assist products used in cars; power management plays
Vicor
(VICR) and Monolithic Power (MPWR); and networking chip plays Marvell and Credo Technology Group (CRDO).

Bolton is also bullish on Taiwan Semiconductor (TSM), “the only practical foundry…to manufacture the most advanced chips such as AI accelerators.” And he sees a boost to the electronic design automation tools sector, including
Cadence Design Systems
(CDNS) and
Synopsys
(SNPS).

In the networking sector, Bolton calls out
Arista Networks
(ANET) as “perhaps one of the primary beneficiaries of AI infrastructure builds.” He thinks the company’s strong position with the large cloud players “gives them a clear advantage in this rapid growth arena.” Needham estimates that AI, over time, will add 5% to 10% to Arista’s long-term growth rate.

Bolton is also bullish on
Cognex
(CGNX), “a world leader in machine vision technology.” He noted that the company has “stepped up the deployment” of deep learning-based technology across its product portfolio.

JMP analyst Andrew Boone had thoughts, too, asserting that the technology will increase the “winner take most” dynamics across the internet. For instance, he thinks the trend should be a boost to social media platforms. “AI models accelerate the brainstorming process for creators, while editing tools are becoming more powerful and easier to use,” Boone wrote. “We think this further democratizes the creative process, which should directly benefit social networks as long-tail content proliferates.”

One of Boone’s conclusions is that the emergence of chatbots like Bard, Bing, and OpenAI is likely to “negatively impact the open web,” with more searches answered by AI. His view is that the ability of AI-based chatbots to answer queries will reduce the number of times information seekers proceed to other websites.

“Simply put this increases the importance of app and direct traffic, or said differently, brand, while likely reducing the business opportunity for the long tail of the web,” he wrote. In particular, Boone thinks the trend “bodes poorly for the long tail of publishers.” He sees potential risks to Taboola (TBLA) and
Outbrain
(OB), which among other things sell ads intended to drive traffic to content publishers’ websites.

Write to Eric J. Savitz at [email protected]