Nvidia 's (NASDAQ: NVDA) transformation from a gaming industry specialist to an artificial intelligence (AI) powerhouse has been nothing short of phenomenal. Accounting for roughly 70% to 95% of the chips used in training and inferencing large language models (according to Mizuho Securities), Nvidia has emerged as the biggest beneficiary and enabler of the ongoing AI revolution.
While Nvidia's stock has gained almost 140% in the past year and a dazzling 471% over the past three years, I don't think the positive impact of this multiyear tailwind on Nvidia's share price will be exhausted in a few quarters. There is still room for the stock to grow -- especially after it became more accessible to shareholders following its stock split .
Nvidia has surpassed consensus revenue estimates in 11 of the past 12 quarters and earnings estimates in 10 of the past 12. While that's no guarantee the company will keep surprising the market, investors awaiting the company's upcoming earnings announcement (scheduled for Aug. 28) can be confident it's taking advantage of a huge opportunity. I think Nvidia stock is still a strong buy for the following reasons.
Dominance in the AI data center market
Nvidia's data center offerings, including graphics processing units (GPUs), AI software stack, supporting frameworks, and other systems, have been in high demand in AI and high-performance computing applications. Cloud service providers, consumer internet companies, and businesses in industries as varied as finance, healthcare, and automotive are relying extensively on the company's solutions to build their AI infrastructure.
Subsequently, the company's data center business saw explosive growth in the first quarter of its fiscal 2025 (ending April 28), with revenue soaring 427% year over year to $22.6 billion. Furthermore, with demand far outstripping supply for the company's next-generation H200 chips and Blackwell architecture chips, Nvidia expects momentum for its data center business to continue in the coming quarters.
Technology leadership
Nvidia's rapid pace of technological innovation has made it difficult for competitors to capture market share. The company has already launched a new GPU architecture called Blackwell, which is 4 times faster in training and 30 times faster in inferencing large language models than the company's flagship H100 GPU. Blackwell architecture has been designed to be backward-compatible with software stack, enabling customers to easily transition to advanced chips. The company has also announced plans to release successors to its H100 GPU annually, a major shift from its previous strategy of launching major chips every two years.
Nvidia has also built a robust software stack to support its AI GPU infrastructure and has partnered extensively with multiple cloud service providers. The recently launched Nvidia Inference Microservices (NIM) software also helps developers quickly build and deploy AI applications. In the face of such developments, any client who wishes to transition to competing hardware has to face high switching costs. Nvidia's commitment to innovation has further resulted in its clients opting to upgrade and expand their AI infrastructure.
Expanding total addressable market
Nvidia has positioned itself as a full-stack AI platform provider to capture a bigger part of the global AI infrastructure spending. Besides advanced AI GPUs, the company also offers all other essential infrastructure required by clients for AI production -- including central processing units, data processing units, advanced networking solutions (InfiniBand, Ethernet, Spectrum-X), software ecosystem (CUDA platform, Nvidia Inference Microservices), and servers.
Nvidia is also gearing up for new data center revenue streams from the sovereign AI market. Governments across the world are striving to develop domestic AI capabilities, and Nvidia is partnering with these national and local players as a major AI infrastructure provider. On the last earnings call, the company said it expects sovereign AI revenue to be in the "high single-digit billions" in fiscal 2025, which will end Jan. 31.
Nvidia expects automotive to be the largest enterprise vertical for its data center business and to generate multibillion-dollar revenue this fiscal year. The company has already supported Tesla in training an AI cluster comprising 35,000 H100 GPUs for autonomous driving. The company has also secured multiple design wins with several leading electric vehicle players for its Nvidia Drive Thor AI car platform.
Improving profitability
Despite ongoing investments in technological innovation, Nvidia has expanded its profit margins. In the first quarter, the company's non-GAAP (adjusted) gross margin jumped by 12.1 percentage points year over year to 78.9%. Net income also soared by 462% year over year to $15.24 billion. This highlights the company's robust pricing power and ability to grow revenue faster than expenses.
Shareholder returns and valuation
Nvidia returned $7.8 billion to shareholders through share repurchases and cash dividends in the first quarter. The company split its stock on a 10-for-1 basis effective June 10, and announced a 150% increase in the quarterly dividend on a post-split basis. These moves highlight Nvidia's confidence in its future growth prospects and commitment to returning value to shareholders.
At recent prices, Nvidia is trading at just under 35 times trailing-12-month sales, a significant premium to its historical five-year average around 26. However, although the stock is expensive, the valuation seems justified considering the solid growth projections for the company. Analysts expect Nvidia's revenue to grow by a solid 97.5% year over year to $120.3 billion, and earnings per share (EPS) to soar by 111.4% year over year to $2.74 in fiscal 2025.
Against this backdrop, investors should consider picking up at least a small stake in this stock ahead of its second-quarter earnings results.
Manali Pradhan has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Nvidia and Tesla. The Motley Fool has a disclosure policy .
Most Popular
Comments / 0