Nvidia or AMD: Goldman Sachs Picks the Superior AI Stock to Buy

Anyone following the stock market knows that AI has been the key driver of the ongoing bull run. But as investors look toward the next phase of AI-driven growth, attention is increasingly shifting from the training of large models to how those models are actually deployed and used in real-world applications. While the build-out of foundational models has driven massive infrastructure investment to date, the long-term opportunity could lie in the rapidly expanding inference market – where models generate results, interact with users, and enable enterprise workflows.

Reflecting this shift in focus, Goldman Sachs analyst James Schneider recently highlighted how the evolving landscape of AI is shaping the future of semiconductor investments.

“We believe AI model training will continue to drive sustainable spending while evolving, potentially with fewer players in the future. However, we think the inference market is set to dramatically increase over the next five years, both from hyperscalers seeking to drive monetization and a much broader swath of corporates seeking to use smaller, cost-effective models to streamline operations and drive incremental sales. We believe this will drive a ‘barbell’ approach to semiconductor investments: (1) performance and software ecosystem leaders; (2) low-cost custom silicon leaders driving greater affordability,” the analyst opined.

With this in mind, Schneider has been assessing the prospects of several AI players, including chip giants Nvidia (NASDAQ:NVDA) and Advanced Micro Devices (NASDAQ:AMD) – believing the ongoing AI opportunity is more pronounced in one than the other. Let’s take a closer look.

Nvidia

There’s no better place to start than with Nvidia, currently the world’s most valuable company – a status it has achieved by becoming the must-own AI stock.

Nvidia’s journey from a gaming-focused chipmaker to the first firm on Wall Street to cross the $4 trillion market cap threshold began when it shifted its focus to data centers. The company recognized early that its GPUs were ideally suited for training deep learning models, and it quickly reoriented its technology toward AI. That pivot placed Nvidia at the heart of the AI boom, with demand surging from hyperscalers, enterprises, and startups alike. Today, it’s no longer just selling chips – it’s selling the core infrastructure behind the AI economy.

Basically, no other company has been able to offer the same level of products across the AI ecosystem that Nvidia can provide. It controls the full stack – from the most advanced AI chips and networking hardware to the CUDA software layer that has become the industry standard. This vertical integration creates a competitive moat that’s proven difficult for rivals to breach, allowing Nvidia to capture an outsized share of both training and inference workloads as AI adoption grows.

That level of excellence has been borne out in consistently strong quarterly reports as was the case in the most recent readout, for F1Q26 (April Quarter). Revenue surged by 69.2% year-over-year to $44.06 billion, beating the Street’s forecast by $810 million. At the bottom line, adj. EPS of $0.81 outpaced expectations by $0.06. And looking ahead to FQ2, even with the guide factoring in an estimated $8 billion hit to H20 revenue tied to recent export restrictions, revenue is anticipated to hit $45 billion, plus or minus 2%, just slightly under the Street’s forecast of $45.66 billion.

Assessing Nvidia’s prospects, while Goldman Sachs’ James Schneider acknowledges that AI infrastructure spending may moderate, the analyst argues that Nvidia is exceptionally well positioned to benefit from the ongoing AI build-out.

“Although we acknowledge that incremental revenue generation potential of Gen-AI technology is not yet clear, we believe there are already early signs (illustrated above) of incremental monetization – and even more compelling signs of cost savings at the enterprise level – which suggest that AI infrastructure CapEx can sustain growth in the next 2-3 years, albeit at slowing rates. As a result, we expect improving visibility for Nvidia’s CY2026 revenue in the coming quarters should help alleviate ‘peak’ concerns and drive a multiple re-rating for the stock,” Schneide wrote.

Accordingly, Schneider rates NVDA shares as a Buy while his $185 price target factors in a one-year gain of 12%. (To watch Schneider’s track record, click here)

That is also the prevalent view amongst Schneider’s colleagues; based on a mix of 37 Buys, 4 Holds, and 1 Sell, the analyst consensus rates the stock a Strong Buy. However, going by the $176.29 average target, a year from now, shares will be changing hands for a ~7% premium — not exactly something to write home about. (See NVDA stock forecast)

AMD

From one chip giant to another – though with a sharply contrasting AI trajectory – AMD’s journey has, so far, been almost a mirror image of Nvidia’s. While Nvidia has surged to the top of the market on the back of its AI dominance, AMD has found it tougher to convince investors that it’s ready to seize the AI opportunity with the same vigor.

A big part of that story comes down to timing and technology. AMD was late to the AI race, lacking both early focus on AI-optimized GPUs and a mature software ecosystem like Nvidia’s CUDA. As a result, AMD has been left playing catch-up in a market its rival had already begun to define and dominate.

Yet, it would be a mistake to count AMD out entirely. The company has a well-established track record of turning the tables on more celebrated competitors. Under Lisa Su’s leadership, AMD transformed from a struggling underdog into a formidable force in CPUs – steadily chipping away at Intel’s dominance by launching superior products and capitalizing on its rival’s missteps.

Now, AMD is working to replicate that success in GPUs, rolling out its Instinct MI series to close the gap with Nvidia’s offerings. Although its ROCm software isn’t as mature as CUDA, it’s improving steadily, and AMD’s commitment to open-source support is attracting developers wary of Nvidia’s closed ecosystem. Moreover, with tech heavyweights like Meta and Microsoft already using Instinct GPUs for AI workloads, AMD is starting to carve out a real presence as a credible Nvidia alternative.

Even if its progress in AI hasn’t yet sparked the same market fireworks, AMD’s broader financials have remained strong. In its latest quarter, 1Q25, AMD reported revenue of $7.44 billion – up 36% year-over-year and $320 million ahead of Wall Street’s forecast. At the other end of the spectrum, adj. EPS of $0.96 beat expectations by $0.03. For Q2, AMD anticipates revenue around $7.4 billion, with a possible variation of $300 million, compared to the consensus estimate of $7.24 billion.

While AMD has been improving its positioning in the AI game, for Goldman’s Schneider, however, the company doesn’t quite have the firepower needed to become a force in the space. 

“AMD has successfully grown its Datacenter GPU market from near zero (excluding HPC) in 2023 to ~$5bn in 2024,” says the analyst. “However, AMD’s performance in the AI market has lagged Nvidia’s on a relative basis driven by high competitive intensity (and software ecosystem investment) from Nvidia and its own hyperscaler customers who are designing custom ASICs with Broadcom, Marvell, and others. We see limited scope for AMD to gain incremental share in this market given Nvidia’s incumbency (particularly among non-traditional and enterprise customers), and as ASICs capture more hyperscaler wallet share for high-volume workloads.”

On that basis, Schneider rates AMD as Neutral, with a $140 price target that implies the stock will remain rangebound for now.

Overall, Schneider is joined by 9 other analysts on the sidelines, while 25 analysts rate AMD a Buy, giving the stock a Moderate Buy consensus. However, the average price target of $135.97 sits about 7% below current levels, suggesting shares may be a bit overvalued – at least for now. It remains to be seen whether the bullish camp revises their targets upward or shifts to a more cautious stance in the months ahead. (See AMD stock forecast)

To find good ideas for stocks trading at attractive valuations, visit TipRanks’ Best Stocks to Buy, a tool that unites all of TipRanks’ equity insights.

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.