rewrite this title Three top Wall Street analysts stay bullish on Nvidia stock. Here’s why

Chip giant Nvidia (NVDA) is considered to be one of the key beneficiaries of the artificial intelligence boom, thanks to robust demand for its advanced graphics processing units (GPUs).

The stock has been under pressure recently due to concerns about valuations of AI plays and growing competition in the AI chip space from rivals like Broadcom (AVGO), Advanced Micro Devices (AMD) and Alphabet-owned Google’s tensor processing units (TPUs). Nvidia is also facing uncertainty related to chip exports to China amid geopolitical tensions between Washington and Beijing.

Despite ongoing pressures, several top analysts remain bullish on Nvidia for several reasons, including its solid track record, strong execution, continued innovation and dominant position in the AI GPU market. TipRanks’ AI Analyst also has an “outperform” rating on NVDA stock with a price target of $205.

Let’s look at the views of three such Wall Street pros who are bullish on Nvidia’s growth potential.

Following a virtual meeting with Nvidia’s vice president of investor relations, Toshiya Hari, Bank of America analyst Vivek Arya reiterated a buy rating on NVDA stock with a price forecast of $275, saying that he continues to view it as a top pick.

Among the key takeaways, Arya mentioned that while Nvidia agrees that Gemini 3 is a top large language model (LLM) that is trained on Google’s in-house TPU, the company contends that it is too early to declare a clear winner. Specifically, the company emphasized that the existing GPU-based LLMs available were all trained on old Hopper (2022) architecture and cannot be compared with the upcoming LLMs that are trained on NVDA’s Blackwell (2024) GPUs.

Arya highlighted that management is confident about the expected launch of the Blackwell-backed LLMs in early 2026, which would prove that “they are at least a full generation ahead of competition.” In fact, external benchmarks like MLPerf and InferenceMAX view Blackwell as the clear leader in both training and inference, with Nvidia standing out in terms of key metrics like tokens per watt and revenue per token.

The five-star analyst added that Nvidia continues to have demand and supply visibility into at least $500 billion of revenue opportunity for Blackwell, Rubin and networking for calendar years 2025 to 2026. Interestingly, the recent deals with ChatGPT maker OpenAI and Anthropic/Microsoft are incremental to this $500 billion outlook (as they are letter of intents) and represent potential upside.

Overall, the meeting reinforced Arya’s bullish thesis, with the analyst finding NVDA stock’s valuation attractive. Specifically, its price-to-earnings (P/E) multiples of 25x and 19x of 2026 and 2027 earnings, respectively, imply only a 0.5x PEG ratio. That’s compared to the average of 2x for the Magnificent Seven stocks and growth competitors.

Arya ranks No. 270 among more than 10,100 analysts tracked by TipRanks. His ratings have been profitable 58% of the time, delivering an average return of 17.7%. 

Bernstein analyst Stacy Rasgon is also upbeat about Nvidia’s prospects and has a buy rating on the semiconductor stock with a price target of $275. In his latest note to investors, the analyst discussed some interesting takeaways from his virtual investor meeting with Stewart Stecker, senior director of investor relations at Nvidia.

Rasgon noted that the $500 billion outlook announced in October for cumulative Blackwell, Rubin and networking sales for calendar years 2025 and 2026 will likely see an upside, as it doesn’t include new deals such as that with Anthropic, the OpenAI 10 GW collaboration, and partnerships in the Middle East.

On concerns about competition from Google’s in-house chips, Rasgon noted that while Nvidia acknowledges the progress that Google has made in more than 10 years, the company believes that it is about two years ahead of the search engine giant’s TPU program.

Nvidia argues that, given the evolving AI market, it will be challenging for Google to persuade cloud service providers to deploy TPUs as they are meant for specific model structures. “But they believe NVIDIA’s programmable platform solutions remain the best hardware for cloud AI infrastructure,” said Rasgon.

Regarding President Donald Trump’s recent post about allowing Nvidia to ship H200 AI chips to China, subject to a 25% cut that goes to the U.S., Rasgon noted that Nvidia is still waiting to secure licenses to ship H200s, after which it intends to look into demand requests and commence manufacturing. Moreover, Nvidia has not yet obtained any details about the 25% revenue sharing with the U.S. government and is currently unclear about how this fee will be accounted for.

Rasgon ranks No. 144 among more than 10,100 analysts tracked by TipRanks. His ratings have been successful 67% of the time, delivering an average return of 27.3%. 

In a research note on the 2026 outlook for semiconductors, Jefferies analyst Blayne Curtis reaffirmed a buy rating on Nvidia stock with a price target of $250. Curtis called Broadcom (AVGO) as the top pick, citing ASIC (application-specific integrated circuits) inflection and the highest level of estimate revisions expected for the company in the semiconductor group. That said, he remains bullish on Nvidia.

“We haven’t given up on NVDA given the technology moat and valuation at 18x the $10 EPS bogey,” said Curtis.

The five-star analyst contends that ASIC adoption is still in its early phases, giving Nvidia plenty of room to grow amid robust spending. He thinks that ongoing worries about NVDA are overstated, given that Blackwell Ultra rollout is on track and Rubin is set to ramp in the second half of 2026.

Furthermore, Curtis noted Nvidia’s dominance in the AI chip space and expects the company’s Vera-Rubin and NVLink 6 launches in the second half of 2026 to bolster its position. He expects Blackwell-backed LLMs to be introduced in the first half of 2026 and act as a potential catalyst for NVDA stock.

Curtis also expects Nvidia’s launch of its new CPX chip in the second half of 2026 to benefit from higher capital spending by hyperscalers and rising focus on inference. The analyst currently expects CPX to generate revenue of $13 billion in calendar year 2027. Based on all these positives, Curtis raised his 2026 and 2027 earnings per share (EPS) estimates for Nvidia to $7.82 and $9.50 from $6.83 and $9.03, respectively.

Curtis ranks No. 58 among more than 10,100 analysts tracked by TipRanks. His ratings have been profitable 64% of the time, delivering an average return of 27.8%.