Bernstein Reaffirms Nvidia and Broadcom Ratings on AI Growth and Valuation Case

Bernstein has renewed its outperform-equivalent ratings on Nvidia and Broadcom, arguing that both companies retain a clear path to AI-driven growth even as investors debate whether demand for advanced computing will cool. The firm set a $400 price target for Nvidia and a $575 target for Broadcom, pairing optimistic forward expectations with an unusual valuation argument: both stocks look inexpensive relative to their own recent histories.

The call is a fresh analyst action, not new financial results from either company. Bernstein analyst Stacy Rasgon’s team said both businesses have visibility into strong growth despite a constrained environment, pointing to tight supply and capacity across much of the hardware needed to build AI systems. Its conclusion depends on that constraint and demand continuing to support customers’ infrastructure spending.

Two routes into AI computing

Nvidia and Broadcom are exposed to that buildout through different products. Nvidia’s graphics processing units, or GPUs, are widely used to train AI models and run complex workloads. Broadcom develops custom chips designed around particular AI workloads; the company’s cited AI customers include Google, Meta, Anthropic and OpenAI.

That distinction helps explain why Bernstein can be positive on both without treating them as interchangeable. Nvidia supplies widely used AI accelerators, while Broadcom’s custom-chip work ties its prospects to customers choosing purpose-built hardware for their own systems. Bernstein’s shared premise is that demand for AI computing remains broad enough to support both approaches.

The forecasts behind the call

Bernstein’s valuation view is inseparable from its growth assumptions. The firm said Nvidia was indicating growth of more than 70% into 2027 from an already large base. It also said Broadcom could potentially double revenue in 2027 and again in 2028 or more. Those are forward-looking expectations, not sales the companies have already recorded.

• For Nvidia, Bernstein said the data-center opportunity remains enormous.
• For Broadcom, it said the company’s AI trajectory appears positioned to accelerate markedly over the next two years.
• The shared constraint is capacity: Bernstein pointed to tight supply across components used in the AI buildout.

Demand confidence meets an unproven outlook

The companies’ leaders have reinforced the demand side of Bernstein’s thesis. Nvidia Chief Executive Jensen Huang and Broadcom Chief Executive Hock Tan said they did not expect AI demand to slow, and Tan said he was not changing his forecasts. Those comments do not settle the question of future spending, but they show management’s view is aligned with the analyst’s premise.

The more consequential test is whether projected demand converts into the growth Bernstein expects. Lower forward multiples may provide room for disappointment, as the firm argued, but they do not remove the underlying dependence on continued AI infrastructure purchases and constrained supply. The ratings are therefore a judgment that the market is underestimating that persistence.