Morgan Stanley has reiterated its bullish stance on Nvidia, maintaining an “overweight” rating and naming the AI chipmaker its “top pick” for 2025. The investment bank issued a price target of $166 per share, representing a 23% upside from Friday’s trading price of approximately $134.82.
The optimism centers on Nvidia’s next-generation Blackwell AI chip, expected to launch in early 2025. Morgan Stanley analysts believe Blackwell will be the “driving force behind revenue” in the second half of next year and could deliver “significant upside for the stock.” CEO Jensen Huang previously described demand for Blackwell as “insane,” sending shares rallying earlier this year.
Addressing Investor Concerns
Morgan Stanley identified four key anxieties that Blackwell’s success could dispel:
1. Slowing Hopper Production: Nvidia’s current-generation chip saw the company forecast just 69.5% revenue growth for Q4, its slowest in seven quarters. However, analysts called this a “non-issue,” noting that Hopper is approaching end-of-life and has a strong backlog of builds already in progress.
2. Staggered Blackwell Rollout: While Nvidia plans to release seven Blackwell variants that won’t ship simultaneously, Morgan Stanley believes timing challenges are temporary. “All of the Blackwells should get sold, even if this drives allocation from one customer to another,” analysts wrote, adding that concerns should “completely disappear” by H2 2025.
3. Competition from Custom Chips: Companies like Broadcom and Marvell produce ASICs (custom AI chips) that compete with Nvidia’s GPUs. However, Morgan Stanley expects “the biggest users of ASICs will actually see purchasing shift back towards GPUs” in 2025, with GPU sales meaningfully outperforming ASIC revenue.
4. Scaling Economics Questions: Some financial backers are questioning returns on expanding GPU cluster sizes. Morgan Stanley countered that Nvidia’s innovations—including its Mellanox acquisition for data center supplies—enhance large cluster efficiency. The analysts noted that inference growth, sovereign training, and enterprise applications represent 70% of data center revenue and provide “multi-year growth drivers.”
Market Context
Despite already rallying 170% in 2024, Wall Street remains broadly bullish on Nvidia heading into 2025, with AI-related investments continuing as a dominant market theme.
Key Quotes
We have tended to be most enthusiastic on NVIDIA when the near-term data points appear mixed, but underlying dynamics are very strong. We think we are approaching that point now.
Morgan Stanley analysts explained their contrarian bullishness on Nvidia despite some mixed near-term signals, suggesting the market may be underestimating the company’s fundamental strength heading into the Blackwell launch.
There are transitional pressures — but by 2h25 the only topic will be the strength of blackwell, in our view.
The investment bank’s analysts predicted that current investor concerns will fade as Blackwell’s success becomes evident in the second half of 2025, making it the dominant narrative for the stock.
All of the Blackwells should get sold, even if this drives allocation from one customer to another, and we think that this persists all year. This is not a concern that should linger.
Morgan Stanley addressed worries about staggered product releases for Blackwell’s seven variants, arguing that demand is so strong that timing issues won’t impact overall sales performance.
Even with any concerns about the AGI arms race cooling down the road, the growth in inference, the growth in sovereign training, the growth in enterprise training applications, and are all multi-year growth drivers, and make up 70% or so of data center revenue.
Analysts highlighted that Nvidia’s growth doesn’t depend solely on the race to artificial general intelligence (AGI), but rather on diverse, sustainable revenue streams that represent the majority of data center business.
Our Take
Morgan Stanley’s analysis reveals a critical tension in AI markets: short-term uncertainty versus long-term conviction. The 23% price target suggests significant upside remains even after Nvidia’s extraordinary 170% gain in 2024, but this optimism hinges entirely on Blackwell’s execution.
What’s particularly noteworthy is the bank’s dismissal of ASIC competition. This contrasts with narratives that hyperscalers would reduce GPU dependence through custom chips. If Morgan Stanley is correct about customers shifting back to GPUs, it validates Nvidia’s platform approach and suggests general-purpose AI accelerators maintain advantages in flexibility and ecosystem support.
The emphasis on inference, sovereign AI, and enterprise applications signals market maturation—moving from pure research and training toward production deployment. This transition could actually expand Nvidia’s addressable market beyond the handful of hyperscalers driving today’s demand. However, the 69.5% growth forecast (though still robust) indicates the hypergrowth phase may be moderating, making execution on Blackwell absolutely critical to maintaining investor enthusiasm.
Why This Matters
This Morgan Stanley analysis underscores Nvidia’s pivotal role in the AI infrastructure boom and highlights how the company’s fortunes are increasingly tied to next-generation chip architectures. The Blackwell launch represents a critical inflection point for the AI industry, as these chips will power the next wave of large language models, AI training, and inference workloads.
The report’s significance extends beyond Nvidia itself—it reflects broader confidence in sustained AI investment despite growing questions about return on investment. Morgan Stanley’s dismissal of competitive threats from custom ASIC chips suggests the market believes general-purpose GPUs will remain dominant for AI workloads, at least through 2025.
For businesses and investors, this analysis indicates that AI infrastructure spending shows no signs of slowing, even as some consolidation may occur. The emphasis on inference, sovereign AI, and enterprise applications signals that AI deployment is moving beyond the “arms race” phase into practical, revenue-generating implementations. This transition could validate the massive capital expenditures hyperscalers have made and sustain the AI boom beyond speculative hype into tangible business value.
Recommended Reading
For those interested in learning more about artificial intelligence, machine learning, and effective AI communication, here are some excellent resources: