It’s no secret that the AI boom of recent years has been a primary driver of market gains, with the introduction of generative and agentic AI models and enormous demand for semiconductor chips and data centers making headlines – and profits for early investors.
Summer Sale – Claim 70% Off TipRanks
But after such a massive run, is it too late to get on board? Top BMO analyst Harsh Kumar, who ranks #5 among more than 10,000 Wall Street analysts, doesn’t think so. In fact, he believes there is still plenty of money to be made in chip stocks.
Kumar notes that the semiconductor sector remains fast-growing, with demand for AI accelerators and memory helping drive the industry toward $1.6 trillion this year.
“AI demand remains strong driven by DC infrastructure deployments and insatiable demand for AI tokens. CSP/hyperscaler and AI lab capex remain elevated in a race to reach AGI. We see AI in the early innings with many years of capex growth to support token demand (+6,050% growth) and tokenomics,” Kumar opined.
And Kumar isn’t stopping with a bullish industry outlook. He’s pointing investors toward two of the biggest names battling for AI chip dollars: Nvidia (NASDAQ:NVDA), the industry’s dominant player, and Advanced Micro Devices (NASDAQ:AMD), its closest rival in AI accelerators. Let’s find out why.
Nvidia
The first of Kumar’s picks that we’ll look at is Nvidia, currently Wall Street’s largest publicly traded company. Nvidia took an early lead in the production of AI-capable GPUs, and has ridden that success to a $5.2 trillion market cap. There are only a dozen trillion-dollar-plus companies trading on Wall Street – but we should note that four of them are semiconductor makers. Nvidia leads the pack in its niche, and the company itself is in the role of the pathfinder.
Nvidia’s fast, high-quality GPU chips, originally developed to meet the graphics needs of high-end gaming, were quickly adopted by the AI industry. They had the speed and processing power necessary to support AI applications, and Nvidia has been regularly releasing updates and new chip models and architectures to keep up with and support advances in the field. The key result of that, for investors, has been the tremendous rise in Nvidia’s share price over the past several years. In the last five years, NVDA is up 885%. The pace of the gains, however, has slowed down recently.
All of this makes Nvidia the dominant player in the AI world today. The company’s semiconductor chips are supporting the industry, and its data center business segment – directly tied to AI – was reported at $75.2 billion in its last set of financial results. In addition, Nvidia’s CUDA platform provides tools, coding guides, and support for the accelerated computing and AI development made possible by the company’s chips and servers, giving Nvidia an industry-leading presence in AI software as well as hardware.
A closer look at Nvidia’s last earnings report, which covered fiscal 1Q27 and the period ending April 26, shows that the company’s total revenue came to $81.6 billion. This was a record level, up 85% year-over-year, and beat the forecast by $2.5 billion. The data center segment, already noted above, was also a quarterly record and was up 92% year-over-year. Nvidia realized a non-GAAP EPS of $1.87, compared to $0.78 in fiscal 1Q26; the EPS was 10 cents per share better than had been expected.
In addition, Nvidia reported a non-GAAP gross margin for the quarter of 75%, and finished the period with $13.24 billion in cash and cash equivalents. That cash reserve was up from $10.61 billion at the end of the previous quarter. In a move that should please investors, Nvidia also announced a significant increase to its quarterly cash dividend – from just $0.01 per common share to $0.25 per share. The new dividend was paid out on June 26, 2026.
Looking ahead, Nvidia’s outlook for fiscal 2Q27 predicts revenue of $91 billion, plus or minus 2%, and expects the non-GAAP gross margin to remain near 75%, within a range of plus or minus 50 basis points. Investors won’t have long to wait to see how those expectations played out, with Nvidia set to report its fiscal second-quarter results after the closing bell on Wednesday, August 26.
BMO’s Harsh Kumar is, unsurprisingly, upbeat on Nvidia, and he sees several reasons for the company’s momentum to continue.
“Nvidia is our top pick as the industry’s lead AI supplier… Demand remains robust with much of NVDA’s capacity sold out for the next 12-plus months. We believe NVDA has a strong financial profile supported by healthy margins and cash flow generation. Looking ahead, we see several growth catalysts including digital factories, autonomous self-driving vehicles, and robotics. We note these opportunities are still several years away. Digital twins and self-driving applications appear closer to deployment than robotics. All three markets are expected to require substantial AI compute, creating additional demand for NVDA’s hardware, software, and AI platform,” the top analyst opined.
Getting into some specifics, Kumar explains why he remains bullish on Nvidia’s near- to mid-term prospects: “Stock trades 18x forward P/E, a discount considering revenue growth of 84% and 50% in FY27 and FY28, respectively, based on our model. We believe market valuations are mispricing an opportunity given defensible full systems end-to-end HW/SW moat.”
Kumar is putting a sizable number behind his bullish Nvidia view. The BMO analyst rates NVDA an Outperform (i.e., Buy) and has a $340 price target on the shares. Reaching that mark would hand investors a return of about 58% over the next 12 months. (To watch Kumar’s track record, click here)
Wall Street isn’t far behind. Nvidia has picked up 26 Buy ratings over the past three months, giving the stock a unanimous Strong Buy consensus. With shares changing hands at $214.72, the $301.82 average price target points to 40.5% upside over the coming year. (See NVDA stock forecast)
AMD
Next on BMO’s radar here is AMD. Like Nvidia above, this mega-cap chip maker has seen its share price make tremendous gains in recent years – for the past five years, the stock is up 340%. The fastest gains have come in the past 12 months; AMD is up 190% since this time last year, and 121% for the year-to-date. AMD now boasts a market cap of $766 billion, and while that figure is hardly in Nvidia’s multi-trillion-dollar league, it still places AMD solidly in the mega-cap category. AMD is now the fourth-largest US semiconductor.
The company’s gains have been driven by its success in building up a strong position as a supplier to the AI industry. AMD’s Instinct line of AI accelerators has become one of the field’s go-to products for supporting AI training and high-speed large language models, while the company’s Helios rack-scale architecture brings MI400-series GPUs together with EPYC CPUs and Pensando networking for large-scale AI deployments.
Just as important as AMD’s success in getting high-performance AI-capable chipsets onto the market is the company’s success in signing deals with major AI model builders. Since October of last year, the company has struck multi-gigawatt GPU deployment agreements and partnerships with OpenAI, Meta Platforms, and Anthropic.
The OpenAI deal, announced last October, calls for the AI developer to deploy up to 6 gigawatts of AMD’s GPUs on a multi-year timeframe, with the first gigawatt to be deployed during 2H26. The Meta agreement calls for AMD to provide up to 6 gigawatts of Instinct GPUs to power Meta’s AI infrastructure. And in the Anthropic agreement, AMD has committed to making a strategic equity investment of up to $5 billion, while Anthropic will deploy up to 2 gigawatts of AMD’s MI450 GPUs in Helios rack-scale solutions, starting in 1H27.
AMD reported its 2Q26 results earlier this month, and showed solid gains at the top and bottom lines. Revenue came in at $11.54 billion, up some 50% year-over-year and beating the forecast by $227.8 million. The company’s non-GAAP EPS of $1.66 was a nickel better than expected, and marked a strong gain from the $0.48 reported for 2Q25. AMD’s data center segment is leading its revenue growth, rising 107% year-over-year to $6.7 billion in Q2, driven by continued strong demand for EPYC processors and Instinct GPUs.
BMO’s Kumar expects AMD’s server strength to continue while the OpenAI, Meta, and Anthropic agreements begin contributing to results, with Helios potentially opening another leg of growth.
“We believe the market is excited about strength in server CPU driven by agentic AI. We believe AMD is on the cusp of remarkable growth led by Helios given the AI deals signed with OpenAI, Meta and Anthropic. Deals should start to bear fruit starting in the second half of 2026/late-September timeframe. We expect operating margins to benefit from expanding top-line growth from AI revenues. We are constructive on gross margin outlook. GM may not benefit as much as OM since the racks system represents a new product class for AMD. We believe it will take several quarters to work through early challenges in manufacturing, supply chain logistics, and deployment. We expect GM to improve gradually from lessons learned and processes mature,” the analyst noted.
Kumar backs up his bullish AMD outlook with an Outperform (i.e., Buy) rating and a $550 price target. From the current share price, his target gives investors a potential 16% return over the next 12 months.
Yet, Kumar is actually on the conservative side compared with Wall Street, where the $647.42 average price target puts the potential one-year gain at about 37%. AMD holds a Strong Buy consensus rating, with 27 of the 33 analysts covering the stock over the past three months recommending a Buy, while the remaining 6 say Hold. (See AMD stock forecast)
Disclaimer: The opinions expressed in this article are solely those of the featured analyst. The content is intended to be used for informational purposes only. It is very important to do your own analysis before making any investment.