For the foreseeable future, major cloud providers, sovereigns, neoclouds, and enterprises are expected to increase their spending on artificial intelligence infrastructure. Cantor Fitzgerald analysts believe that this technology is “the only meaningful growth driver” for chip stocks, and they have predictions about which ones will gain the most?
In a statement released on Wednesday, analysts at Cantor stated that they anticipate these factors “to continue driving the AI trade” as tech giants increase capital expenditures and point to the need to invest more in AI infrastructure, while AI hardware businesses exhibit “strong product cycles.”
Nvidia Corp., which is currently expanding its Blackwell AI platform, was selected as the best choice by the Cantor team. The Blackwell ramp, according to the analysts, positions Nvidia “for meaningful beats/raises” and may boost its profits per share to $8 in the upcoming year, supporting their $240 price target. Additionally, that would be much higher than the $6.31 average projection for Nvidia’s EPS in the upcoming fiscal year.
Cantor also anticipates that exposure to AI would increase the stocks of Advanced Micro Devices Inc., Broadcom Inc., Micron Technology Inc., and Taiwan Semiconductor Manufacturing Company Ltd.
According to the researchers, AI is now “an area of relative certainty within a very uncertain world” due to geopolitical tensions and economic difficulties. For instance, TSMC told that the U.S. government has withdrawn its permission for exporting certain chip-making technology and equipment to its factory in China, and that it would no longer have verified end-user status on December 31. Additionally, SK Hynix Inc. and Samsung Electronics Co. had their VEU status withdrawn.
However, the experts stated that although momentum stocks have lost some of their shine, this is really a short-term issue. According to the experts, recent stories claiming organizations have struggled to integrate AI into their operations are “overdone.” They don’t care since they believe that the hyperscale cloud companies’ return on capital is “still strong.”
Capital expenditures at Microsoft Corp., Meta Platforms Inc., Alphabet Inc., and Amazon.com Inc. are expected to rise 57% this year and 20% in 2026, according to analysts, up from 40% and 9% two months ago.
The Cantor team also stated that the AI trade is experiencing “some agita” as a result of reports that the Chinese government is discouraging local companies from adopting US-made technology, specifically Nvidia’s H20 chip, which it is preparing to sell to customers in China again after being essentially banned from doing so by the Trump administration in April.
Cantor, on the other hand, sees the risks “as noise today,” and the analysts are digging down on their “bullish thesis” that AI development and implementation are still in their infancy.
According to the Cantor analysts, there is increasing hope for AMD’s data-center graphics processing units as the firm prepares for its analyst day in November. Moreover, AMD’s client and server central processing unit businesses have “only accelerated higher” due to higher average sales prices and share gains, as well as a “intact” outlook for its Instinct AI accelerator series. Earlier this year, investors were worried that customers were making purchases in advance to beat out potential tariff price increases later on. For the year, experts estimate AMD’s earnings per share to be at $4. $3.85 is the consensus of FactSet.
AMD’s push to offer rack-scale solutions is one factor that gives analysts “high hopes around the company’s ability to significantly grow its penetration across the AI landscape,” even though the chip maker’s data-center revenues are predicted to remain “relatively small” at an estimated $6.5 billion this year. AMD’s “vision of capturing a non-trivial share in large scale training clusters” and the increasing need for AI inference were also mentioned.
