NVIDIA Q2 Revenue Exceeds $96 Billion, Targets 70% Growth by 2028

Original | Odaily Planet Daily (@OdailyChina)

Author | Azuma (@azuma_eth)

At 1:00 AM Beijing Time on August 27, NVIDIA announced its second-quarter financial results for fiscal year 2027, ending July 26, 2026. The financial data showed that NVIDIA’s (NVDA) revenue and profits for the second quarter significantly exceeded expectations, with its data center business continuing to accelerate, and guidance for the next quarter setting a new record.

Specifically, NVIDIA's Q2 revenue reached $96.221 billion, representing a 106% year-over-year increase and an 18% sequential growth, surpassing analyst expectations by 4%. Non-GAAP net profit amounted to $53.954 billion, up 118% year-over-year, with adjusted EPS at $2.22, nearly 6% above market expectations. Additionally, NVIDIA provided a Q3 revenue guidance of $108 billion, signaling the company’s imminent entry into the era of "quarterly revenue exceeding $100 billion."

The above figures are already impressive, but they are not the most astonishing number in this earnings report. During the investor call following the earnings release, NVIDIA CFO Colette Kress provided a signal that the market had not fully anticipated—the company expects revenue growth of approximately 70% in fiscal year 2028 (excluding any revenue from data centers in China), significantly exceeding Wall Street’s previous expectation of 45%.

More importantly, NVIDIA emphasized that this 70% growth forecast was made even under conditions of supply constraints. In other words, the primary issue the company now faces appears to have shifted from “How long will AI demand last?” to “How much of NVIDIA’s demand can the supply chain actually meet?”

Following its earnings report and conference call, NVIDIA rose over 4% in after-hours trading, currently at $219.53.

Looking closely at NVIDIA's $96.2 billion in revenue, it becomes clear that data center business remains the absolute core source of income.

In Q2, NVIDIA's data center revenue reached $89 billion, representing a 117% year-over-year increase and an 18% quarter-over-quarter increase, accounting for over 90% of the company's total revenue. More importantly, the growth rate of data center revenue continues to outpace the company's overall revenue growth—In Q1, NVIDIA's total revenue grew 85% year-over-year, while data center revenue grew 92%; by Q2, these growth rates further accelerated to 106% and 117%, respectively. Even as quarterly revenue approaches $100 billion, NVIDIA's core business has not shown any significant slowdown—on the contrary, it is accelerating further.

From an internal perspective of the data center business, Q2 growth continued to be primarily driven by "hyperscalers." NVIDIA disclosed that revenue related to hyperscale customers in Q2 amounted to approximately $48.7 billion. Meanwhile, major tech companies such as Amazon, Microsoft, Google, and Meta continue to expand their investments in AI infrastructure. Just prior to the earnings release, AWS announced plans to further deploy 2 million NVIDIA GPUs.

This means that, at least based on the current revenue structure, the primary driver of AI computing demand remains the capital expenditures of major cloud providers. This is the most critical variable for observing NVIDIA’s future growth trajectory and the continuation of the AI cycle: if capital spending by companies like Microsoft, Amazon, Google, and Meta continues to rise, NVIDIA’s high growth will still have direct order support; conversely, if these companies’ AI investments enter a consolidation phase, NVIDIA will be the first to feel the pressure.

On the other hand, NVIDIA’s edge computing business generated $7.2 billion in revenue, representing a 27% year-over-year increase and a 13% sequential growth. While its revenue scale still lags significantly behind that of the data center business, it serves as a key representative of NVIDIA’s strategic push into “physical AI” and reflects its long-term effort to reduce dependence on a single business segment.

The most significant product-level signal in this earnings report and subsequent earnings call is the launch of Vera Rubin—NVIDIA confirmed on the call that Vera Rubin has entered full-scale production and began shipping earlier this month. Each gigawatt of Vera Rubin deployed represents approximately $40 billion in revenue opportunity, and it is expected to contribute about 20% of data center business revenue in Q3.

This announcement marks NVIDIA's gradual transition from the Blackwell cycle to the Rubin cycle.

For a company like NVIDIA, which has an extremely rapid product iteration cycle, this transition between old and new platforms is crucial. In the past, a major concern in the market regarding AI chip cycles was whether each new generation could successfully pick up the baton; if there was a noticeable gap between two generations, revenue growth would naturally decline sharply. This time, Rubin’s commercialization pace is clearly fast enough.

As NVIDIA’s rack-scale supercomputing platform designed specifically for AI, Vera Rubin integrates seven proprietary chips—including the Rubin GPU, Vera CPU, NVLink6, and ConnectX-9—delivering a transformative leap in performance. This evolution ultimately translates into enhanced value capture: the benchmark of “1 GW of compute power → $40 billion in revenue” indicates that, as data center scale continues to expand following Vera Rubin’s deployment, NVIDIA will capture increasing value from each newly built AI data center.

In addition, Jensen Huang himself emphasized on the earnings call that the scaling speed of Vera Rubin will set the fastest record in the company’s history... this may be the confidence behind NVIDIA’s assertion that revenue will still grow by approximately 70% in fiscal year 2028.

Regarding the question of whether demand for AI has peaked, NVIDIA’s management conveyed a stronger signal than ever during the post-earnings investor call, indicating that demand is not only not slowing down but continues to accelerate.

Jensen Huang explicitly stated on the earnings call that AI has entered a new phase. Over the past year, AI compute demand was primarily driven by a few leading model companies, but now demand is spreading to a broader range of areas, including more cutting-edge models, open-source models, enterprise AI, agents, and robotics.

The most significant indicator of change is the accelerating growth in inference demand. As large models transition from training to real-world applications, AI compute consumption is no longer a one-time investment during training but continues to rise with increased user usage and token generation. Jensen Huang has even directly described this shift by saying, “Tokens are becoming productive and profitable.”

In addition, NVIDIA stated on the earnings call that there is no sign of noticeable slowdown in demand based on current capital expenditure plans by major tech companies. The world’s five largest cloud providers are expected to spend nearly $800 billion on capital expenditures this year, potentially rising to around $1.3 trillion next year.

These observations collectively form the foundation of NVIDIA’s assessment of future AI demand. However, under such strong demand, supply has become the limiting factor constraining NVIDIA’s further growth. In other words, the company’s primary bottleneck is not a lack of orders, but whether components, production capacity, and delivery capabilities can keep pace with the speed of orders.

NVIDIA's management explicitly stated on the earnings call that supply will remain the primary constraint on the company's growth through at least fiscal year 2028. Jensen Huang emphasized that, without supply constraints, the company's fiscal year 2028 outlook would be "much higher."

Here, "supply" is not limited to GPUs alone. As AI chip scales expand rapidly, HBM, advanced packaging, wafers, networking equipment, and data center power may all become critical constraints on the speed of compute deployment. In particular, HBM: NVIDIA directly addressed the significant rise in memory prices during this earnings call and expects its non-GAAP gross margin for Q3 to be approximately 74%, down from 75% in Q2. The ongoing increase in memory costs is putting pressure on future profit margins.

To overcome this bottleneck, NVIDIA is expanding its role to become a "capital organizer" for AI infrastructure. Financial disclosures reveal that NVIDIA has established strategic partnerships with Apollo, BlackRock, Blackstone, Brookfield, Goldman Sachs, and KKR to mobilize over $500 billion in third-party capital for AI infrastructure development in the coming years. Meanwhile, NVIDIA is collaborating with SB Energy on the PORTS-Pike Technology Campus in Ohio to secure land, power, and building shell capacity in advance to support the deployment of its computing resources.

NVIDIA’s earnings report, along with management’s comments during the earnings call, has temporarily answered the market’s most pressing question once again—whether AI demand has peaked and the cycle has ended.

Looking at NVIDIA’s guidance of 70% revenue growth for fiscal year 2028, the company’s current message remains: “The cycle is far from over”—the ceiling is still far away, and today’s growth rate isn’t even a true reflection of demand, but merely the limit of what the supply chain can support. When a company with quarterly revenues nearing $100 billion still dares to set growth targets far exceeding market expectations despite supply constraints, it can only mean that the orders NVIDIA is seeing are even stronger than the numbers it feels comfortable disclosing.

Meanwhile, NVIDIA itself is adapting to market demands by continuously expanding its business role. Today, NVIDIA is no longer just a semiconductor company designing GPUs—it is evolving into a general contractor for “AI factories,” providing chips, networking, CPUs, software, deployment solutions, and even participating in infrastructure financing and securing land and power supply.

Of course, high expectations come with high pressure. The modest downward revision of gross margin guidance and policy uncertainties in global markets are inevitable friction points in scaling to a trillion-dollar valuation... but at least for now, the core logic behind the market’s willingness to grant NVIDIA a premium remains intact—on the赛道 of AI infrastructure, NVIDIA remains the rule-maker, with no visible competitors in sight.

When NVIDIA's Q3 revenue officially surpasses the $100 billion mark, the company will enter a new order of magnitude. The 70% growth guidance for next year suggests that this threshold may merely be the starting point for the next phase.