Morgan Stanley reinstates this AI chipmaker as its top pick in semiconductors
Investing.com -- Morgan Stanley has reinstated Nvidia as its top pick in the semiconductor sector, citing the company’s strong positioning amid data center constraints and the rise of agentic artificial intelligence applications.
Following a recent road show with Nvidia CEO Jensen Huang, CFO Colette Kress, and Head of Investor Relations Toshiya Hari, Morgan Stanley analysts highlighted several factors supporting their renewed conviction in the stock. The firm believes Nvidia’s advantages in addressing land, power, and data center shell constraints, along with component financing capabilities, position the company favorably in the current market environment.
Nvidia Corporation
Morgan Stanley reinstated Nvidia as its top semiconductor pick, noting the stock trades at an undemanding valuation of 15 times fiscal year 2028 earnings. The firm emphasized several key strengths following investor meetings with company leadership.
The analysts highlighted Nvidia’s diverse customer base beyond hyperscalers and frontier labs. Management noted that half of the business, likely the faster-growing portion, comes from a broader range of AI model companies, neoclouds, sovereign entities, and enterprise customers. This diversity helps mitigate land, power, and shell constraints affecting U.S.-based hyperscalers.
Morgan Stanley pointed to Nvidia’s strategy of increasing revenue per gigawatt substantially. The company showed that its Feynman architecture, expected in 2028, will drive revenue per gigawatt from $40 billion to well above $50 billion. The Vera Rubin platform is expected to deliver up to 25 times token improvement versus Blackwell.
The firm also addressed the impact of agentic AI on demand. While persistent agents will drive higher CPU demand, Nvidia management indicated that GPU spending will remain the largest portion. The company projects an agentic user may generate 10 times the daily tokens of a human user.
Morgan Stanley noted Nvidia has mobilized half a trillion dollars of non-Nvidia capital to invest in AI data centers. The company expects to grow CPU revenue from $20 billion in 2026 to $40 billion in 2027.