Morgan Stanley Names Nvidia Top Semiconductor Pick
Morgan Stanley has named Nvidia its top semiconductor pick, citing an undemanding valuation and the potential of its upcoming 2028 Feynman architecture.
By Muhamed Porić
October 8, 2026 at 3:41 PM

Morgan Stanley has reinstated Nvidia (NVDA) as its top semiconductor pick. The firm cites an "undemanding" valuation and the company’s ability to navigate shifting bottlenecks in AI infrastructure development. The upgrade reflects a change in how analysts view the firm's growth trajectory relative to the AI capital expenditure cycle.
"We certainly see NVIDIA as well positioned, with a very undemanding valuation," said Joseph Moore, analyst at Morgan Stanley, in a recent note to clients.
Valuation and Earnings Outlook
Nvidia’s current market valuation is a primary driver for the firm's renewed optimism. According to Morgan Stanley’s projections, the stock is trading at 15 times the firm's estimated earnings for fiscal 2028. This multiple is being weighed against a projected shift in the semiconductor sector, where the primary constraints for AI growth are moving away from raw chip production toward the physical speed of constructing and financing data center capacity.
As of October 8, 2026, Nvidia shares were trading at $236.14. Morgan Stanley analysts suggest that Nvidia’s strategic positioning allows it to benefit from these infrastructure constraints, as the company remains central to the hardware requirements of large-scale data center operators.
The 'Feynman' Architecture Impact
Beyond current market conditions, the bank’s outlook hinges on the expected performance of Nvidia’s upcoming "Feynman" architecture, slated for 2028. Analysts anticipate that this architecture will improve the economic efficiency of AI deployments. The firm projects that Feynman will increase revenue per gigawatt from the current $40 billion to more than $50 billion.
What Is at Stake for AI Infrastructure
This shift in analyst sentiment highlights the evolution of the AI hardware market. Early phases of the AI boom were defined by the scarcity of high-end GPUs. The market is now entering a phase where the ability to integrate power, space, and cooling, measured by revenue per gigawatt, becomes the critical metric for measuring long-term profitability. By focusing on the efficiency of its 2028 architecture, Morgan Stanley suggests that Nvidia is moving to capture value from the secondary phase of the data center build-out.
Muhamed Porić
Founder and Editor of Embers.
Newsletter
Get Embers in your inbox
The stories that actually moved something, delivered when there's something worth sending, not daily filler.