After soaring more than 800% over the last two years, shares of Nvidia (NVDA -0.12%) have experienced a downturn in early 2025 following China’s DeepSeek announcement that it developed an advanced large language model at a low cost. This raised questions about the demand for Nvidia’s graphics processing units (GPUs), which account for over 80% of its total revenue.
Investment growth from key customers
In response to the DeepSeek news, Nvidia’s top customers have reported earnings that indicate a rising demand for AI chips. Cloud service providers generated approximately half of Nvidia’s $30 billion in data center revenue for the fiscal third quarter ending in October. Amazon, as the leading cloud provider, is significantly investing in technology, with capital expenditures increasing to $83 billion in 2024 and plans to reach around $100 billion in 2025, primarily directed toward technology infrastructure for cloud and AI services. Amazon utilizes Nvidia’s H100 GPU for generative AI workloads within Amazon Web Services and is expected to adopt Nvidia’s Blackwell computing platform launching this year.
Meta Platforms is another major Nvidia customer that aims to boost its capital spending, with plans to raise its budget to between $60 billion and $65 billion in 2025, up from $37 billion last year. Meta has been deploying Nvidia’s H100 GPUs extensively, training its Llama 4 models on a large cluster of 100,000 GPUs. Meta also projects substantial long-term investments in AI infrastructure, indicating a strong outlook for Nvidia, which holds an estimated 70% to 95% market share of AI chips.
Diversification beyond data centers
Nvidia’s data center revenue growth has overshadowed other significant business opportunities. The company’s gaming segment, for instance, is recovering, with sales of gaming GPUs increasing from $1.5 billion in the third quarter of fiscal 2023 to $3.2 billion in the same quarter of 2025, representing an annual run rate exceeding $12 billion.
Additionally, Nvidia achieved record revenue in its automotive segment, which includes the Drive computing platform for training autonomous vehicles. Automotive revenue rose from $251 million in the third quarter of 2023 to $449 million in the same quarter two years later. As the adoption of autonomous driving technology increases, this segment could develop into a multibillion-dollar opportunity.
Nvidia has previously stated that its total addressable market is $1 trillion, with the potential for growth as the company explores new markets for its technology. At the J.P. Morgan Healthcare Conference in January, Nvidia highlighted that the rising number of interactions in health services will create new avenues in robotics and AI solutions where its GPUs could play a crucial role.
The emergence of DeepSeek’s cost-effective large language model is unlikely to impede Nvidia’s long-term prospects, as it may drive faster innovation in AI, favoring the leading AI chip supplier.
The recent decline in Nvidia’s stock has reduced its valuation to a forward price-to-earnings multiple of 30, which analysts consider a potential bargain given expectations of a 51% revenue and earnings increase this year.
Moreover, Microsoft recently unveiled a new chip, suggesting that quantum computing is becoming achievable in the near term rather than the distant future, alongside predictions from Google and IBM about imminent changes in computing technology. Quantum computing is expected to facilitate calculations that would take current systems millions of years, potentially accelerating advancements in fields such as medicine and chemistry. As these developments unfold, Nvidia is positioned to play a key role in integrating AI and quantum computing technologies.
Disclaimer: The content of this article is for informational purposes only and should not be construed as investment advice. We do not endorse any specific investment strategies or make recommendations regarding the purchase or sale of any securities.
Featured image credit: Nvidia
