AI demand is accelerating! NVIDIA Corporation (NVDA.US) reported a revenue increase of 100% year-on-year in Q2, with guidance for a 70% growth rate in the next fiscal year far exceeding expectations, and has reached an agreement with Amazon.com, Inc. for 2 million GPUs.

Global AI chip leader NVIDIA Corporation (NVDA.US) delivered a quarterly performance report on Wednesday that is sure to be recorded in historyrevenue of $96.22 billion, a year-over-year increase of 106%, net profit of $59.69 billion, and adjusted earnings per share of $2.22, with all three core metrics significantly exceeding Wall Street's expectations. Even more exciting for the market was the company's first-ever long-term guidance predicting approximately 70% revenue growth in fiscal year 2028, far surpassing analysts' prior expectations of around 45%. However, this "textbook-quality" financial report initially saw a nearly 3% drop in after-hours trading, followed by a rapid recovery of over 4% after the earnings call, creating a V-shaped reversal. Given that the stock price has declined the day after five of the last six earnings reports, the market has come to treat "surpassing expectations" as a default scenarioonly guidance strong enough to warrant an upward revision of future earnings forecasts can reignite valuation expansion. Data Overview: The "Inertia of Surpassing Expectations" at $96.2 Billion For the second fiscal quarter ending July 26, 2026, NVIDIA Corporation's core metrics all substantially exceeded market expectations: Revenue was approximately 5.7% higher than the company's prior guidance of $91 billion, marking the largest overshoot in nearly two years. Free cash flow reached $21.3 billion, with adjusted net profit soaring 118% year-over-year. Third-Quarter Guidance: Meeting the "Whisper Number" of $108 Billion For the third quarter, NVIDIA Corporation expects revenue of $108 billion (2%), exceeding the market expectation of $104.2 billion. This guidance also assumes no inclusion of any revenue from data center computing in China. Purchase commitments surged from $119 billion in the previous quarter to $279 billion, primarily related to memory procurementan indication of demand as well as a source of cost pressure. The CFO specifically noted that shipments of Hopper data center products to China accounted for less than 1% of data center revenue this quarter. As analysts expect NVIDIA Corporation's profit margins to face pressure due to the ramp-up of Rubin chip production and the rising memory prices in the silicon supply chain, the company's profit margins are under close scrutiny. In terms of gross margin, NVIDIA Corporation forecasts an adjusted gross margin of 74% (50 basis points) for the third quarter, down from 75% in the current quarter. The CFO warned that due to skyrocketing memory costs, the gross margin is expected to bottom out in the fourth fiscal quarter (ending January next year) at 71% to 72%, stabilizing between 72% and 73% for fiscal year 2028. Data Center: The "Absolute Engine" with $89 Billion The data center business remains the core pillar of NVIDIA Corporation's growth, with revenue reaching $89 billion, a year-over-year increase of 117%, accounting for 92.7% of total revenue. In terms of customer structure, NVIDIA Corporation adjusted its data center business disclosure last quarter, categorizing customers into two groups: Hyperscale and ACIE (AI cloud, industrial, and enterprise customers). In this quarter: Hyperscale customers: Revenue of $48.71 billion, a year-over-year increase of 102%, and a quarter-over-quarter increase of 13%. ACIE customers: Revenue of $40.31 billion, a year-over-year increase of 138%, and a quarter-over-quarter increase of 25%. The growth rate of ACIE business has surpassed that of hyperscale customers, indicating that the demand for AI computing power is spreading from leading cloud providers to sovereign AI, regional cloud service providers, and a wider range of enterprise customers. NVIDIA Corporation has been committed to expanding its customer base, demonstrating that its sales are no longer as reliant on a few tech giants as before. CEO Jensen Huang stated in the earnings report: A year ago, it was driven by a single lab; today, we are entering a golden age of new AI labs and startups, multiple leading labs are scaling in parallel, the open model ecosystem is thriving, and physical AI is coming online. Fiscal Year 2028 Guidance: 70% Growth Overwhelms Expectations, Supply Remains a Bottleneck NVIDIA Corporation provided its first revenue growth outlook for fiscal year 2028approximately 70%. This number far exceeds the market's previous estimate of around 45%. Huang stated during the earnings call that he had never given guidance a year ahead of time before. CFO Kress clearly pointed out that supply will continue to be a bottleneck limiting growth at least until fiscal year 2028. Huang further mentioned that without supply constraints, the company's performance outlook for fiscal year 2028 would be much higher. Surge in Expenditure by Large Tech Companies in AI NVIDIA Corporation is one of the most valuable companies globally, and its performance is seen as a barometer for the AI market, as its chips power most major data centers and advanced AI models worldwide. Weeks before the report, several companies, including Microsoft Corporation and Metatwo of NVIDIA Corporation's major clientsreiterated their earlier expectation that large tech companies plan to invest over $730 billion in artificial intelligence infrastructure this year, a significant increase from last year's $400 billion. As a key chip manufacturer at the core of the AI boom, NVIDIA Corporation holds an optimistic sales outlook for fiscal year 2028, alleviating concerns that AI spending might lose momentum. NVIDIA Corporation is the highest-valued company globally and a leading supplier of AI accelerators, the crucial components for training and running AI models. This position ensures that NVIDIA Corporation's quarterly earnings serve as a bellwether for the entire industry. The CFO of NVIDIA Corporation also mentioned that if the company could secure more supply, its growth rate would be even faster. "Incredible as it may seem, even at our current scale, demand is accelerating," she said. "Customer forecasts indicate that our growth will double next year." Major Collaboration with AWS: Additional Deployment of 2 Million GPUs The earnings report also announced that NVIDIA Corporation has expanded its cooperation with Amazon.com, Inc. AWS. Kress revealed that the two parties will deploy an additional 2 million NVIDIA Corporation GPUs in Amazon.com's global infrastructure in 2027 and 2028. This collaboration further solidifies NVIDIA Corporation's dominance among hyperscale cloud service providers. Huang pointed out that market demand is "accelerating," and AWS's additional orders are the latest testament to this trend. Vera Rubin Fully Online: A New Engine for 20% of Data Center Revenue In NVIDIA Corporation's second-quarter earnings report, Huang stated that market demand is accelerating. He also emphasized the launch of the companys latest chip series, Vera Rubin. "AI infrastructure building is proceeding at full speed," he said. "Vera Rubin is now fully in production, designed specifically to support this moment." While Blackwell is still being shipped in large quantities, NVIDIA Corporation's next-generation AI system, Vera Rubin, has fully entered mass production. CFO Colette Kress revealed on the earnings call that Vera Rubin began shipping earlier this month. For every 1 gigawatt of Vera Rubin computing power deployed, there corresponds approximately $40 billion in revenue opportunity. Vera Rubin is expected to contribute about 20% to data center business revenue in the third quarter. Vera Rubin is no longer a single chip but rather a complete POD-level system, consisting of Vera CPU, Rubin GPU, Groq 3 LPX, BlueField-4 storage, and Spectrum-6 network. NVIDIA Corporation's supply chain covers over 350 factories across 30 countries, with system vendors like Dell, HPE, Lenovo, and Supermicro all entering mass production. Companies like CoreWeave, Google Cloud, Microsoft Azure, Oracle, and Nebius are also deploying related systems. Huang stated: Vera Rubin has achieved full mass production, designed for this moment. Concerns Over Circular Financing: The "Gray Area" of Hundreds of Billions of Dollars Beyond the exceeding expectations figures, the market's most focused variable remains NVIDIA Corporation's increasingly complex circular financing arrangements. On one hand, NVIDIA Corporation is participating in a $500 billion AI infrastructure financing platformpartnering with six financial institutions including Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs Group, Inc., and KKR, aiming to mobilize over $500 billion in third-party capital over the long term. On the other hand, the company may provide credit support of up to $105 billion for data center projects rented by OpenAI. Goldman Sachs Group, Inc. is also voicing concernsthe market's worries about "circular financing" are growing, and NVIDIA Corporation must provide detailed explanations during the earnings call. The core controversy surrounding these arrangements is whether they are releasing genuine and financially constrained demand for computing power or whether they are using NVIDIA Corporation's own and partnering financial institutions' credit support to pull some future demand into the present. NVIDIA Corporation's logic is both offensive (funding disruptors to accelerate global AI deployment) and defensive (funding non-hyperscale ecosystems to diversify its customer base). However, the critical question remainsif the revenue data for AI inference does not grow enough to cover the high costs of computing power, these financing arrangements may shift from being enablers to sources of risk. To this end, in this earnings report, NVIDIA Corporation listed debt as an independent risk factor for the first time, warning that increasing external financing commitments could have an adverse effect on financial status and cash flow. As of July 26, the company's outstanding senior notes amounted to $33.5 billion, with a commercial paper program size of $25 billion. The company expects to pay off $15 billion in debt over the next one to five years, while the figure in the previous quarter was only $2.75 billion. NVIDIA Corporation warned: Maintaining our debt, contractual restrictions, and additional debt issuance may result in a significant portion of our operating cash flow being used for debt repayment and principal repayment. The company stated that while such financing deals could accelerate AI adoption and create greater product demand, critics worry that circular financing may fuel artificial demand. Shareholder Returns: $26 Billion Distributed, $99 Billion Awaiting In terms of capital return, NVIDIA Corporation returned approximately $26 billion to shareholders this quarter through stock buybacks and dividends. As of the end of the quarter, the remaining stock buyback authorization was about $99 billion. The cash dividend for the next quarter is set at $0.25 per share. The company also issued $25 billion in senior unsecured notes for general corporate purposes. Increasingly Fierce Competitive Threats However, an increasing portion of tech companies' planned expenditures is shifting toward in-house chip development, aimed at reducing reliance on NVIDIA Corporation's costly and supply-constrained processors. At the same time, a host of potential competitors are eyeing the lucrative market that NVIDIA Corporation dominates. Furthermore, NVIDIA Corporation's customers are increasingly developing their own chips, which could reduce their reliance on NVIDIA Corporation in the long run. As artificial intelligence is used more frequently for automating tasks and answering queries, NVIDIA Corporation's graphics processors are facing intensifying competition from central processing units and custom chips better suited for what is known as inference processes. This shift is prompting major tech companies to invest in developing their own chips. Just this week, OpenAI, the maker of ChatGPT, announced that its new Jalapeno processor outperformed NVIDIA Corporation's current processor lineup in testing. Reports indicate that Meta plans to start production of its custom "Iris" AI chip in September, which is part of its fourth-generation custom chip initiative aimed at reducing computing costs. According to reports in June, Alphabet has ordered over 3 million chips from Intel Corporation, expected to be delivered in 2028. The report also noted that NVIDIA Corporation is assessing the manufacturing technology of this American chipmaker to produce a processor that integrates four GPUs into a single unit. Competitors Intel Corporation and AMD are also targeting the inference market, with several Chinese companies, including Baidu Inc Sponsored ADR Class A, already producing chips for such tasks. NVIDIA Corporation has noticed this trend. In March of this year, NVIDIA Corporation released a new type of central processor and AI system based on licensed technology from the inference-focused startup Groq. This deal, valued at $17 billion, combines Groq's chips with its upcoming Vera Rubin platform. Earlier this month, SpaceX CEO Elon Musk announced that the company will exclusively use NVIDIA Corporation's hardware, further solidifying NVIDIA Corporation's leading position in the AI market. NVIDIA Corporation stated that by 2027, its AI chips could have a revenue potential exceeding $1 trillion, doubling from the $500 billion predicted for its Blackwell and Rubin chips before 2026.