$2.8B in Daily Revenue, But Market Value Plummets by Trillions

NVIDIA, one of the cornerstone companies in the technology sector, recently faced a setback as its stock price dipped over 5% after the company reported its guidance for the fourth quarter, which fell short of peak expectations. Following the release of their third-quarter earnings report for the fiscal year 2025, the market reacted sharply, reflecting the sensitivity of investors to performance forecasts in the fiercely competitive tech landscape.

On November 21, during early hours in Beijing, NVIDIA unveiled its financial results, boasting an impressive revenue of $35.1 billion for the third quarter, marking a staggering year-on-year growth of 93.7%. This accounted for an average daily earnings rate of approximately $389 million. Additionally, the company reported a net profit of $19.31 billion, demonstrating a remarkable increase of 108.9%. Notably, NVIDIA exceeded analysts' expectations, who had anticipated revenues of $33.25 billion and profits of $16.93 billion.

The company’s gross margin stood at 74.6%, a slight uptick of 0.6% compared to the previous year. CFO Colette Kress attributed this margin improvement primarily to the increased sales of AI chips designed for data centers. In the third quarter of the fiscal year 2025, NVIDIA’s data center business generated a record $30.8 billion, which represents a 112% growth year-on-year and a 17% increase quarter-on-quarter. This business unit is crucial, given the surging demand for high-performance computing to support artificial intelligence (AI) applications

Advertisement

During the conference call following the earnings report, NVIDIA's CEO Jensen Huang shared insights on the latest developments concerning the next-generation AI graphics processor, Blackwell. Huang announced that production rates for Blackwell chips have significantly surpassed expectations. Notably, major partnerships have emerged, with companies like Microsoft, Oracle, and OpenAI beginning their uptake of these advanced chips. CFO Kress confirmed that around 13,000 chip samples have already been dispatched to customers, illustrating the urgency and excitement surrounding this newer technology.

Despite the strong quarterly results, NVIDIA lowered its expectations for the fourth quarter, projecting revenues of $37.5 billion. This forecast indicates a deceleration in revenue growth from 94% to 69.5%, stirring investor anxiety and leading to the aforementioned drop in stock price after the announcement, which equated to a staggering $179 billion loss in market capitalization.

The slowing revenue momentum raises questions about potential technical bottlenecks related to the new chip generation. NVIDIA's operations are categorized mainly into four sectors: data centers, gaming and AI PCs, professional visualization, and automotive and robotics.

The data center segment, which comprises GPU chips and AI servers, recorded revenue of $30.8 billion in the third quarter, showcasing a remarkable year-on-year growth of 112%. However, compared to the previous quarter's growth rate of 154%, this data indicated a reduction in pace. The increase in data center revenues can be credited to rising GPU demand spurred by supercomputing facilities and NVIDIA's broader alliances with major players like AWS, Lenovo, and Ericsson.

In the gaming and AI PC sphere, revenue climbed to $3.3 billion in the same quarter, marking a 15% increase year-on-year and a 14% quarterly rise. The surge in this sector is bolstered by increased demand for GPUs in both personal computers and gaming consoles. NVIDIA has also rolled out 20 new games leveraging its GeForce RTX technology and DLSS (Deep Learning Super Sampling) during this quarter, with manufacturers like ASUS and MSI releasing new AI PCs powered by RTX.

Professional visualization achieved a revenue of $486 million, a 17% year-on-year and a 7% quarter-on-quarter increase. In their report, NVIDIA highlighted ambitions through partnerships such as the collaboration with Foxconn, which is utilizing NVIDIA's Omniverse platform for building digital twins and industrial AI to accelerate the production of NVIDIA's GB200 Grace Blackwell superchips.

The automotive and robotics sector has also reported rapid growth, with revenues of $449 million for the third quarter, reflecting increases of 72% year-on-year and 30% quarter-on-quarter, fueled largely by rising sales of autonomous driving and robotics chips.

Despite impressive profits, NVIDIA has observed a deceleration in revenue growth and earnings per share (EPS) projection this quarter. While revenue growth soared by 94%, it follows a trio of extraordinarily high growth rates over the past three quarters of 122%, 262%, and 265%. Notably, NVIDIA's non-GAAP adjusted EPS stood at $0.81, exhibiting a robust year-on-year growth of 103%, albeit lower than the previously reported growth rate of 152% in the preceding quarter.

However, NVIDIA faces scrutiny over potential risks related to its new Blackwell AI chips. Reports from The Information highlighted concerns about severe overheating issues affecting the performance of these chips within high-capacity server racks, raising questions about their reliability and necessitating design modifications and enhanced cooling systems.

In the wake of AI hype, instigated by the popularization of ChatGPT, NVIDIA has emerged as a significant player on the stock market. By June 2023, the company's valuation surged past the $1 trillion mark, and by June 2024, it crossed $2 trillion. By November of the same year, NVIDIA solidified its position as the company with the highest market capitalization globally, amassing a valuation of $3.65 trillion, surpassing even Apple’s former peak.

The ongoing race for computational power continues to elevate NVIDIA GPUs as indispensable tools for enterprises looking to leverage AI technologies. Big tech firms and unicorn startups have ventured into this landscape, intensifying competition—not only for superior algorithms and data but also for securing NVIDIA chips as critical resources.

In early 2024, Meta's CEO Mark Zuckerberg declared his company’s intentions to procure 350,000 H100 GPU chips from NVIDIA, aiming to bolster their AI infrastructure by the end of that year. The cost projection for this acquisition hovers between $8.75 billion and $10.5 billion, considering the individual price point of the H100 chip ranges from $25,000 to $30,000.

As NVIDIA's core clientele, the "big seven" in the U.S. stock market, which also includes Apple, Microsoft, Google, Amazon, Meta, and Tesla, have all released their third-quarter financial reports for fiscal year 2025, consistently expressing continued investment in foundational AI infrastructure—AI GPUs.

Notably, NVIDIA has eclipsed traditional rivals like Intel and AMD, which recorded revenues of $13.3 billion and $6.819 billion, respectively, coupled with losses of $16.6 billion and profits of $771 million. Research has indicated that NVIDIA commands between 80% to 90% of the market share for AI training and inference chips.

Recognizing the high costs associated with AI chips, many tech titans acquire NVIDIA chips while concurrently engaging in their own self-development of AI hardware. Amazon, for instance, has developed its own AI chips, Inferentia and Trainium. During promotional periods, Amazon strategically deployed 80,000 AI chips for cloud computing, improving performance by about 40% to 50% compared to NVIDIA’s chips, ultimately allowing them to halve operational costs for similar tasks.

Despite these emerging threats, experts argue NVIDIA remains irreplaceable in the short term. The ecological advantage that NVIDIA has established with its CUDA architecture—an initiative launched in 2006 to simplify GPU programming—remains a pivotal reason for the company’s dominance. CUDA’s advantages have created a self-propagating cycle where more developers using the architecture only reinforce its usability, further attracting new talents.

Considering that the CUDA platform is intrinsically tied to NVIDIA hardware, developers seeking alternate platforms encounter new risks and cost assessments similar to mobile developers contemplating a departure from the Android ecosystem. The very root of NVIDIA's current success revolves around its ability to foster an expansive developer community, which ultimately feeds back into the stability and prestige of its products.