BofA’s Vivek Arya Sees Nvidia at $350 as Agentic AI Drives an “Unprecedented” Chip Cycle

Photo of Omor Ibne Ehsan
By Omor Ibne Ehsan Published

Quick Read

  • Nvidia (NVDA) trades around $215 with Bank of America’s $350 price target based on 85% revenue growth, 75% gross margins, and 50-60% expected earnings growth, with the company sitting on $119B in supply commitments and raising its dividend by nearly 25x. AMD (AMD) is up 109% year to date on strong agentic AI demand, Broadcom (AVGO) reported $8.4B in AI semiconductor revenue up 106% year over year with guidance to $10.7B next quarter, and Lam Research (LRCX) posted record $5.84B revenue as supply constraints benefit the entire semiconductor bill of materials.

     

  • Agentic AI applications consuming exponentially more tokens and GPU compute than chatbots are creating unprecedented semiconductor demand that analysts say has shifted from a cyclical chip market to a permanent structural shift in pricing and capacity requirements.

     

  • The analyst who called NVIDIA in 2010 just named his top 10 AI stocks. Get them here FREE.

BofA’s Vivek Arya Sees Nvidia at $350 as Agentic AI Drives an “Unprecedented” Chip Cycle

© Apichatn / Shutterstock.com

Bank of America’s Vivek Arya walked onto CNBC with a number that sits well above consensus. His price target on NVIDIA (NASDAQ:NVDA | NVDA Price Prediction) is $350, against a stock trading around $215.

Arya is making a longer call, and this might lead to other analysts also revising their estimates upwards in the coming days and weeks.

From chatbots to agents, and why the compute curve bends

Arya’s setup: “What we are seeing is kind of this unprecedented wave of demand for semiconductors because of generative AI. What was simply chat bots have now moved to the next stage, which is agentic applications. These are multi-step autonomous applications.”

A chatbot answers a question once. An agent decomposes a task, calls tools, retries, reasons across steps, and runs in the background. Each step consumes tokens, and tokens consume GPUs.

Jensen Huang made the same case on the most recent earnings call, telling investors “Agentic AI has arrived, doing productive work, generating real value and scaling rapidly across companies and industries” and describing the data center buildout as “the largest infrastructure expansion in human history.”

The capex-to-growth correlation

Arya’s second observation should make CFOs uncomfortable about underspending. “The more you have invested in your computing infrastructure, the faster you’re growing. You see that kind of straight line correlation.” If true, capex restraint becomes a competitive liability, which is why hyperscaler orders keep stepping up rather than rationalizing.

NVIDIA’s balance sheet reflects that confidence on the supply side. The company is sitting on $119.0 billion in supply-related commitments, which only makes sense if demand keeps compounding.

Arya also says the exceptional spending has led to a “real generational squeeze on the supply side.”

Unpacking the $350 target

Arya’s math is straightforward. “Our price target is $350. How many other multi-trillion dollar market cap companies are showing 85% sales growth? We saw very strong margins. We saw them raise their dividend by a factor of almost 25 from before.”

Q1 FY27 revenue came in at $81.61 billion, up 85.23% year over year, beating consensus by 3.16%. Non-GAAP gross margin held at 75.0%. The quarterly dividend went from $0.01 to $0.25 per share, alongside an $80 billion additional buyback authorization. The full Q1 FY27 8-K lays out the details.

Arya’s punchline: “Even if there is no multiple expansion, I think what we will see is just earnings growth, which could be 50 to 60%, that is capable of driving very strong returns for this company.” At a forward PE of 24x, the bull case hinges on earnings continuing as they are, not on a re-rating.

The supply squeeze spreads

Unprecedented demand creates unprecedented scarcity, which is where the rest of the complex earns its multiples. AMD (NASDAQ:AMD) is up 109% year to date, with CEO Lisa Su citing “strong momentum as inferencing and agentic AI drive increasing demand for high-performance CPUs and accelerators.”

Broadcom (NASDAQ:AVGO) reported AI semiconductor revenue of $8.40 billion, up 106% year over year, with Hock Tan guiding to $10.7 billion in Q2 AI semis. Lam Research (NASDAQ:LRCX) posted record revenue of $5.84 billion, and Marvell is now 73% data center, with custom AI design activity CEO Matt Murphy described as at an “all-time high.”

Arya stretches the point to the whole bill of materials. “If you could make anything in semiconductors, whether it was wafers or substrates or memory or lasers or anything, you did extremely well because of the supply constraint.”

Where Intel fits, and where it does not yet

Intel is up 204% year to date on the foundry turnaround narrative. Arya thinks Intel “can become part of that solution”, but the company needs 2 to 4 years to catch up to Taiwan Semiconductor on manufacturing. Intel’s Xeon 6 was selected as the host CPU for NVIDIA’s DGX Rubin NVL8 systems, a design win that monetizes only if the foundry roadmap holds.

Watch into August earnings: whether the $91.0 billion Q2 revenue guide holds, what happens to gross margin as Blackwell 300 ramps, and whether China data center contribution returns. If Arya is right, all three move the same direction.

 

Photo of Omor Ibne Ehsan
About the Author Omor Ibne Ehsan →

Omor Ibne Ehsan is a writer at 24/7 Wall St. He is a self-taught investor with a focus on growth and cyclical stocks that have strong fundamentals, value, and long-term potential. He also has an interest in high-risk, high-reward investments such as cryptocurrencies and penny stocks.

Continue Reading

Top Gaining Stocks

DELL Vol: 15,291,396
HP
HPQ Vol: 48,674,188
NTAP Vol: 6,668,169
SWKS Vol: 5,338,626
EL Vol: 8,107,759

Top Losing Stocks

CTRA Vol: 73,319,495
COIN Vol: 7,927,507
TTWO Vol: 7,048,109
UHS Vol: 1,236,515
CHTR Vol: 2,101,059