Earnings Preview | "The Most Important Stock on Earth" NVIDIA's Results Are About to Be Announced! Will "AI Faith" Create Another Huge Wave?

According to the Zhitong Finance APP, NVIDIA (NVDA.US), the world's highest market capitalization company and referred to by Goldman Sachs as "the most important stock on Earth," will announce its third-quarter earnings for the fiscal year 2025, ending in October, after the U.S. stock market closes on Wednesday, which is Thursday morning Beijing time. Following a 240% surge in stock price in 2023, NVIDIA's stock price has nearly tripled this year, fueled by the global "AI faith." Global investors are looking forward to NVIDIA once again announcing unparalleled results that far exceed market expectations, reigniting confidence in artificial intelligence and driving the global stock market into an "AI carnival."

Wall Street analysts generally expect this highest market capitalization company to report continued explosive revenue growth, thanks to the surging demand from global enterprises and core government departments for artificial intelligence (AI) infrastructure to expand or build new data centers. Whether it is the U.S. tech giants Meta, Microsoft, Google, and Amazon, which have reported significant increases in spending on the core AI infrastructure—NVIDIA AI GPUs—or AI application leaders like OpenAI, xAI, Anthropic, and Palantir, or major government departments focusing on "sovereign AI systems" continuing to invest heavily in building AI data centers, there are no signs of a slowdown in the global enthusiasm for artificial intelligence.

According to the latest analyst expectation data compiled by Visible Alpha, Wall Street analysts generally expect this chip giant's total revenue for the third quarter to grow by 84% year-on-year, reaching $33.28 billion, with net profit soaring to $17.45 billion, or $0.70 per share, far exceeding the $9.24 billion from the same period last year, which was $0.37 per share. This also means that NVIDIA will achieve record revenue and net profit for six consecutive quarters.

NVIDIA underwent a 1-for-10 stock split in June, and as of last Friday's U.S. stock market close, NVIDIA's stock price reached $141.98. Ahead of the earnings release, Wall Street is extremely bullish on NVIDIA's stock price trajectory over the next 12 months. Among them, Wall Street banking giant Bank of America recently reaffirmed its "buy" rating on NVIDIA and significantly raised its target price from $165 to $190, above the general target price on Wall Street.

The "burning money wave" for artificial intelligence among global companies, including the "Big Six" such as Meta, Microsoft, and Google, remains unstoppable, and the stock logic closely related to AI chips is extremely robust. The stock price of AI chip leaders like NVIDIA may not have stopped rising yet. Especially for NVIDIA, which holds an 80%-90% market share in the data center AI chip sector, its stock price may continue to set historical highs, and breaking through the generally expected $163 by Wall Street analysts may just be a matter of time

The wave of global data center spending shows no signs of stopping! Are the giants continuing to burn cash, with NVIDIA emerging as the biggest winner?

According to expected data from Visible Alpha, Wall Street analysts anticipate that NVIDIA's data center business total revenue for the third fiscal quarter is expected to rise to another historical high of $29.53 billion, following a record high of $26.3 billion in the second quarter.

In August, NVIDIA CEO Jensen Huang stated, "Global data centers are fully committed to modernizing the entire computing stack through accelerated computing models based on NVIDIA AI GPUs and increasing efforts to deploy generative AI."

Morgan Stanley, a major Wall Street firm, earlier this week raised NVIDIA's stock 12-month target price from $150 to $160 and maintained its "overweight" rating, with the core logic being that the growth of the data center business is expected to be even stronger. Morgan Stanley's analysis team stated, "We expect NVIDIA's data center business to drive most of the revenue growth contribution over the next five years, as the market's enthusiasm for generative AI creates a very favorable environment for AI/(machine learning) hardware parallel computing solutions."

Currently, among the seven major tech giants in the U.S. stock market (the Magnificent Seven), besides AI chip leader NVIDIA (NVDA.US), the other "six giants" have already released their earnings reports. Overall, their earnings reports, along with those of TSMC, the "king of chip foundries," basically show that the demand for AI GPUs continues to exhibit explosive growth, indicating that NVIDIA is the biggest winner of this earnings season in the U.S. stock market.

These tech giants are increasing spending to build artificial intelligence data centers to meet the nearly endless demand for AI inference/training computing power. As tech giants make comments about continuing to burn cash around increasing artificial intelligence capital expenditures, Wall Street investment firm D.A. Davidson stated that NVIDIA will continue to thrive and remain one of the biggest beneficiaries of the global AI boom for some time to come. "For example, as long as Microsoft continues to spend heavily to upgrade its massive data centers, we believe it will continue to transfer shareholder value from Microsoft to NVIDIA," the D.A. Davidson research team stated.

When discussing the market demand for AI chips, TSMC's leader C.C. Wei mentioned during the earnings conference that the outlook for AI chip demand is very optimistic and emphasized that TSMC's customers' demand for CoWoS advanced packaging far exceeds the company's supply. "The company will fully respond to customers' demand for CoWoS advanced packaging capacity. Even if capacity doubles this year and continues to double next year, it will still be far from enough." TSMC's CoWoS S/L/R advanced packaging capacity is crucial for the production capacity of NVIDIA's Blackwell AI GPUs and other broader AI chips According to the latest forecast data from Wall Street financial giant Citigroup, by 2025, the capital expenditures related to data centers of the four largest technology giants in the United States are expected to grow by at least 40% year-on-year. These massive capital expenditures are largely tied to generative artificial intelligence, indicating that the computing power demand for AI applications like ChatGPT remains substantial. Citigroup stated that this means the giants are likely to continue to significantly expand their spending on data centers beyond the already strong spending levels in 2024. The firm expects this trend to provide a substantial positive catalyst for the stock prices of NVIDIA, the undisputed leader in AI GPUs, and data center interconnect (DCI) technology providers.

The four technology giants referred to in Citigroup's research report are global cloud computing leaders Amazon, Google, Microsoft, along with the parent company of social media platforms Facebook and Instagram. In this latest research report, Citigroup predicts that by 2025, the capital expenditures of these four tech giants on data centers will grow by 40% to 50% year-on-year. The substantial increase in spending by tech giants on data centers is expected to continuously attract international capital to the stock prices of NVIDIA, Arista Networks, and other data center network technology giants, which are considered "shovel sellers" in the global AI field.

UBS Wealth Management predicts that this year, large tech companies' annual spending on artificial intelligence will grow by 50%, reaching $222 billion, and then increase by another 20% in 2025. "Microsoft, Alphabet, Amazon, and Meta account for nearly half of all AI spending, and their strong balance sheets and investment willingness may continue to support robust growth in AI spending." UBS stated in a report, "Investors should take advantage of recent volatility to build sufficient exposure to high-quality AI stocks."

Analysts' Key Focus: How is Blackwell's shipment volume? Has the overheating issue been resolved?

Investors will also focus on the shipment outlook data update for NVIDIA's latest AI GPU, which is based on the Blackwell architecture. Jensen Huang described Blackwell architecture products as "disruptors of the AI industry." The Blackwell architecture AI GPU made its debut in the MLPerf Inference v4.1 benchmark test, showcasing its outstanding performance in large language model (LLM) inference tasks. Specifically, in the inference task of the open-source large model Llama 2 70B, the performance of the Blackwell architecture AI GPU per GPU is approximately 4 times better than that of the H100.

Additionally, reports have indicated that the latest generation flagship AI GPU, the Blackwell architecture AI GPU, has serious overheating issues in high-capacity server racks. These problems have led to design adjustments and project delays, causing major clients like Google, Meta, and Microsoft to worry about whether they can deploy Blackwell servers as planned. Wall Street analysts will inquire in the earnings conference call whether Jensen Huang has resolved the overheating issues of AI server racks with the Blackwell architecture AI GPU and the shipment outlook for this year and early next year Recently, analysts from Morgan Stanley warned that some supply-side constraints may limit the upside potential of NVIDIA's recent performance outlook, adding, "We believe that larger upward adjustments will occur later this year." NVIDIA stated in its last earnings call that it expects to achieve "billions of dollars" in revenue from Blackwell architecture AI GPUs in the fiscal quarter ending in January as production increases.

Regarding the overheating issues of AI server racks, media reports cited sources revealing that the Blackwell architecture AI GPUs are designed specifically for AI and HPC, but overheating problems have occurred in servers configured with 72 processors, with power consumption per rack reaching up to 120 kilowatts. Reports indicated that the overheating issues forced NVIDIA to modify the rack design multiple times, which not only limited GPU performance but could also damage hardware. Major clients are concerned that these technical issues may delay the deployment of processors in data centers.

According to media reports, to address this challenge, NVIDIA has requested suppliers to adjust the rack design and is working with partners to optimize the cooling system. Although such engineering improvements are a routine step in large-scale technology releases, they have further delayed product delivery timelines.

It is understood that a spokesperson for NVIDIA responded to this matter by stating that the company is closely collaborating with cloud service providers, and design adjustments are part of the normal R&D process. NVIDIA hopes that through this collaboration, the final products will meet expected performance metrics and reliability while accelerating the resolution of technical bottlenecks.

The Blackwell architecture AI GPUs, carefully refined in collaboration with TSMC, began limited production at the end of October this year for trial use by NVIDIA's core customers and to configure their respective data centers, with mass production and shipment expected as early as the end of January next year