Key Points
Nvidia stock is back in the spotlight as three top Wall Street analysts stick with bullish calls, even while the shares face fresh pressure from valuation worries, rising competition, and uncertainty tied to chip exports to China.
The core debate is straightforward: Nvidia is widely seen as a major beneficiary of the artificial intelligence boom because of strong demand for its advanced graphics processing units. But with investors questioning how much optimism is already priced in—and rivals pushing harder into AI chips—some market watchers have been looking for signs that Nvidia’s growth story is losing momentum.
These analysts are arguing the opposite. Across recent notes and investor meetings, they point to Nvidia’s execution, continued innovation, and what they describe as a dominant position in the AI GPU market. TipRanks’ AI Analyst also rates NVDA stock “outperform,” with a $205 price target.
Why Wall Street is staying bullish on Nvidia stock
The recent pressure on Nvidia stock has been tied to a mix of concerns that keep showing up in investor conversations:
- Valuation anxiety around AI-related stocks
- Intensifying competition in AI chips from rivals such as Broadcom, Advanced Micro Devices, and Alphabet-owned Google’s tensor processing units
- Geopolitical uncertainty tied to exports to China, amid tensions between Washington and Beijing
Even with those overhangs, the analysts highlighted here stay constructive on Nvidia stock for a consistent set of reasons: a strong track record, confidence in management’s execution, and belief that upcoming product cycles—especially Blackwell—can reinforce Nvidia’s lead.
A key theme is timing. Several analysts leaned on the view that comparisons between today’s leading large language models and what comes next may be premature, especially if much of the existing GPU-trained model landscape is still tied to older architecture.
Nvidia stock: Bank of America’s Vivek Arya sees Blackwell as a generation ahead
Bank of America analyst Vivek Arya reiterated a buy rating on Nvidia stock with a $275 price forecast after a virtual meeting with Nvidia’s vice president of investor relations, Toshiya Hari. Arya said he continues to see the company as a top pick.
One of Arya’s central takeaways was Nvidia’s stance on Google’s progress with in-house chips. Nvidia agreed that Gemini 3 is a top large language model trained on Google’s TPU, but argued it is still too early to declare a clear winner in the broader AI infrastructure race.
The company’s reasoning, as relayed by Arya, focused on the training hardware behind many current models. Nvidia emphasized that existing GPU-based large language models were trained on the Hopper (2022) architecture, and the company believes those results should not be directly compared with upcoming models trained on Nvidia’s Blackwell (2024) GPUs.
Arya pointed to management’s confidence in the expected launch of Blackwell-backed large language models in early 2026. In Nvidia’s view, that timing could help demonstrate that the company remains “at least a full generation ahead of competition.”
Arya also cited external benchmarks referenced in the discussion, including MLPerf and InferenceMAX, which he said view Blackwell as a clear leader in both training and inference. The benchmarks highlighted metrics such as tokens per watt and revenue per token, where Nvidia believes it stands out.
Beyond performance, Arya’s bullish view leaned heavily on scale. He said Nvidia has demand and supply visibility into at least $500 billion of revenue opportunity spanning Blackwell, Rubin, and networking across calendar years 2025 and 2026.
He added that recent deals with ChatGPT maker OpenAI and Anthropic/Microsoft were described as incremental to that $500 billion outlook because they are letters of intent—meaning they represent potential upside rather than being included in the existing figure.
On valuation, Arya argued Nvidia stock looks attractive versus peers. He pointed to price-to-earnings multiples of 25x and 19x for 2026 and 2027 earnings, respectively, which he said implies a 0.5x PEG ratio. He compared that to an average of 2x for the Magnificent Seven stocks and growth competitors.
Nvidia stock: Bernstein’s Stacy Rasgon flags upside beyond the $500 billion outlook
Bernstein analyst Stacy Rasgon also maintained a buy rating on Nvidia stock and set a $275 price target, sharing takeaways from a virtual investor meeting with Stewart Stecker, Nvidia’s senior director of investor relations.
Rasgon focused on Nvidia’s previously announced $500 billion outlook, introduced in October, for cumulative Blackwell, Rubin, and networking sales across calendar years 2025 and 2026. His view: that number may prove conservative.
He said the $500 billion outlook likely has upside because it does not include newer deals and initiatives referenced in discussion—such as a partnership with Anthropic, the OpenAI 10 GW collaboration, and partnerships in the Middle East.
On competitive pressure from Google’s TPU efforts, Rasgon noted Nvidia’s acknowledgment that Google has made meaningful progress over more than a decade. Still, Nvidia’s position, as described by Rasgon, is that it remains about two years ahead of Google’s TPU program.
Rasgon said Nvidia also believes the fast-moving nature of the AI market makes it difficult for Google to persuade cloud providers to deploy TPUs broadly, since the chips are designed for specific model structures. In contrast, Nvidia argues its programmable platform approach remains the best hardware foundation for cloud AI infrastructure.
Rasgon also addressed uncertainty tied to China shipments. He referenced a recent post by President Donald Trump about allowing Nvidia to ship H200 AI chips to China, subject to a 25% cut that goes to the U.S.
According to Rasgon, Nvidia is still waiting to secure licenses to ship H200 chips. After receiving those approvals, the company plans to assess demand requests and begin manufacturing. Rasgon added that Nvidia has not yet received details about the proposed 25% revenue-sharing arrangement, and it remains unclear how such a fee would be accounted for.
The takeaway from Rasgon’s note was less about a single headline and more about unresolved variables: licensing, demand visibility after licensing, and the mechanics of a proposed revenue share. For investors tracking Nvidia stock, those items may shape expectations around international sales execution.
Nvidia stock: Jefferies’ Blayne Curtis stays positive as Blackwell Ultra and Rubin roll out
Jefferies analyst Blayne Curtis reaffirmed a buy rating on Nvidia stock with a $250 price target in a research note focused on the 2026 semiconductor outlook.
Curtis named Broadcom as Jefferies’ top pick, pointing to an ASIC inflection and what he sees as the strongest estimate revisions across the semiconductor group. Even so, he made clear he has not turned negative on Nvidia stock.
Curtis argued that Nvidia’s technology moat and its valuation—18x the “$10 EPS bogey,” as he described it—keep him constructive.
A major point in Curtis’ view is that ASIC adoption remains early, leaving room for Nvidia to keep expanding even as custom chips become a larger part of the AI infrastructure discussion. He said worries about Nvidia may be overstated, particularly with the Blackwell Ultra rollout on track and Rubin expected to ramp in the second half of 2026.
Curtis also pointed to additional product and platform catalysts he expects to support Nvidia’s position, including Vera-Rubin and NVLink 6 launches in the second half of 2026.
He expects Blackwell-backed large language models to arrive in the first half of 2026, and he framed that timing as a potential catalyst for Nvidia stock.
Curtis also highlighted Nvidia’s planned CPX chip in the second half of 2026, tying the opportunity to higher capital spending by hyperscalers and increasing focus on inference. In his estimates, he expects CPX to generate $13 billion in revenue in calendar year 2027.
Reflecting his more upbeat outlook, Curtis raised his earnings per share estimates for Nvidia to $7.82 for 2026 and $9.50 for 2027, up from $6.83 and $9.03, respectively.
What investors are watching next
For Nvidia stock, the bullish case in these notes depends less on broad AI excitement and more on a sequence of product and execution milestones that analysts believe can reset the narrative.
Based on the points raised by the analysts:
- Blackwell-backed large language models expected to launch in early 2026 or the first half of 2026
- Continued rollout progress, including Blackwell Ultra on track and Rubin ramping in the second half of 2026
- How licensing and policy details develop for potential H200 shipments to China, including unanswered questions around the proposed 25% revenue sharing with the U.S.
- Whether Nvidia’s performance claims—supported by benchmarks such as MLPerf and InferenceMAX in analyst commentary—continue to reinforce its lead in training and inference
At the same time, the competitive backdrop remains central. The analysts repeatedly returned to the idea that challengers are real, but that Nvidia’s programmable platform, pace of innovation, and ecosystem strength keep the company in a favorable position.
Where the bullish calls land
Despite the near-term pressure points around valuation, competition, and geopolitics, the three analysts highlighted here kept their positive stances:
- Bank of America’s Vivek Arya: Buy, $275 price forecast
- Bernstein’s Stacy Rasgon: Buy, $275 price target
- Jefferies’ Blayne Curtis: Buy, $250 price target
Adding to the broader tone, TipRanks’ AI Analyst rates NVDA stock “outperform” with a $205 price target.
Taken together, the message is clear: these analysts see Nvidia stock as a name that can keep compounding if the Blackwell and Rubin cycles land as expected and if demand stays strong across the company’s targeted opportunity set.
Conclusion
Nvidia stock is navigating a more skeptical market moment, with investors weighing AI valuations, intensifying chip competition, and unresolved questions around China-related exports. Still, top-ranked analysts remain firmly bullish, pointing to Nvidia’s execution track record, upcoming Blackwell-backed large language models expected in early 2026, and a massive revenue opportunity tied to Blackwell, Rubin, and networking across 2025 and 2026.
For now, the analysts’ shared view is that Nvidia’s product cadence—and proof points in training and inference performance—could do the most to settle the debate, especially as the next generation of models begins to roll out on newer hardware.
