Nvidia (NVDA 2.63%) stock is currently down 12% from its all-time high. It suffered a sharp sell-off in January after China-based start-up DeepSeek asserted that it had trained a competitive artificial intelligence (AI) model using a fraction of the computing power that had been deployed by leading U.S.-based developers like OpenAI.
Investors feared that DeepSeek’s techniques would be adopted by other AI developers, leading to a substantial drop in demand for Nvidia’s high-end graphics processing units (GPUs), which are the best hardware available for developing AI models. However, those concerns might have been overblown.
Google parent Alphabet (GOOG -0.54%) (GOOGL -0.49%) is a big buyer of Nvidia’s AI data center chips, and on Feb. 4, its CEO, Sundar Pichai, made some comments that should make Nvidia’s investors feel much better.
Image source: Nvidia.
The DeepSeek saga
DeepSeek was established in 2023 by a successful Chinese hedge fund called High-Flyer, which had been using AI to build trading algorithms for years. DeepSeek released its V3 large language model (LLM) in December 2024, followed by its R1 reasoning model in January, and their competitiveness with some of the latest models from OpenAI and other start-ups got the tech sector buzzing.
Since DeepSeek’s work is open source, the industry quickly learned some important details. The start-up claims to have trained V3 for just $5.6 million (not including an estimated $500 million in chips and infrastructure, according to SemiAnalysis), which is a drop in the bucket compared to the tens of billions of dollars spent by companies like OpenAI to reach their current stage of development.
DeepSeek also used older generations of Nvidia’s GPUs like the H100, because the U.S. government banned the chip maker from selling its latest hardware to Chinese firms (to protect America’s AI leadership).
It turns out DeepSeek implemented some unique innovations on the software side to make up for the lack of computational power. It developed highly efficient algorithms and data input methods, and it also used a technique called distillation, which involves using the knowledge from an already-successful large AI model to train a smaller model.
In fact, OpenAI has accused DeepSeek of using its GPT-4o models to train DeepSeek R1, by prompting the ChatGPT chatbot at scale to “learn” from its outputs. Distillation rapidly speeds up the training process because the developer doesn’t have to collect or process mountains of data. As a result, it also requires far less computing power, which means fewer GPUs.
Naturally, investors are worried that if every other AI developer adopted this approach, it would trigger a collapse in demand for Nvidia’s chips.
Nvidia is gearing up for a record year of GPU sales
On Feb. 26, Nvidia will report its financial results for its fiscal 2025, which ended Jan. 31. The company expects to have generated $128.6 billion in total revenue, which would be a whopping 112% increase from the previous year. Recent quarterly results suggest around 88% of that revenue will be attributable to its data center segment, thanks to soaring GPU sales.
According to Wall Street’s consensus forecast (provided by Yahoo), Nvidia could set another record in its current fiscal year 2026, with $196 billion in total revenue potentially in the cards. Hitting that estimate will depend on further demand for GPUs from AI developers, so it’s easy to understand why investors are nervous over the DeepSeek news.
While the H100 remains a hot product, Nvidia’s latest GB200 GPU — which is based on its Blackwell architecture — can perform AI inference at up to 30 times the speed. Inference is the process by which an AI model absorbs live data (like a chatbot prompt) and produces an output for the user. It typically comes after the initial training phase (more on this in a moment).
The GB200 is currently the gold standard for AI data centers, and demand was significantly outstripping supply when it started shipping to customers at the end of 2024.
Image source: Alphabet.
Sundar Pichai counters
Pichai held a conference call with Wall Street analysts on Feb. 4 to discuss Alphabet’s 2024 fourth quarter results. In response to one of their questions, he said there has been a notable shift in the allocation of computing power over the last three years, with a growing amount going toward inference compared to training.
Pichai said newer reasoning models (like DeepSeek’s R1 and Alphabet’s Flash Thinking models) will only accelerate that shift. These models spend more time “thinking” before producing a response, so they require significantly more computing power than their predecessors. The technical term for that is test-time scaling, and it’s a way for AI models to deliver more accurate information without conducting more pre-trailing scaling (which involves feeding models endless amounts of new data).
Meta Platforms CEO Mark Zuckerberg has similar thoughts. He recently said a drop in training workloads won’t necessarily mean developers need fewer chips, because capacity is simply shifting toward inference instead.
Finally, Alphabet told Wall Street it plans to allocate $75 billion to capital expenditures (capex) during 2025, most of which will go toward data center infrastructure and chips. That figure represents a significant increase from its 2024 capex of $52 billion, so the company certainly isn’t pulling back.
All in all, it appears the demand picture for Nvidia’s GPUs is still very much intact. Considering its stock is trading at an attractive valuation right now, the recent dip might even be a buying opportunity.
Randi Zuckerberg, a former director of market development and spokeswoman for Facebook and sister to Meta Platforms CEO Mark Zuckerberg, is a member of The Motley Fool’s board of directors. Suzanne Frey, an executive at Alphabet, is a member of The Motley Fool’s board of directors. Anthony Di Pizio has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Alphabet, Meta Platforms, and Nvidia. The Motley Fool has a disclosure policy.