NVIDIA's 70% Revenue Growth Forecast: AI Chip Heat Wave – Continuation or Peak?

One-sentence conclusion: Understanding NVIDIA's 70% revenue growth forecast allows prediction of TSMC's capital expenditure direction for the next two years.

NVIDIA announced its Q2 2026 financial results, showing year-over-year revenue growth of 70%, far exceeding market expectations. This figure not only reflects the robust demand for AI chips but has become a key indicator for judging whether the AI boom can continue. However, high base effects, US export controls on China, and competitor catch-up have led the market to question the growth momentum for subsequent quarters. This article analyzes the significance of this 70% growth from three angles: technological drivers, market impact, and investment advice.

Angle One: Core Technology/Event Interpretation – Dissecting the Driving Forces Behind the 70% Forecast

NVIDIA's 70% year-over-year revenue growth stems from three layers: First, explosive growth in data center GPU demand. With the proliferation of large language models (LLMs) and generative AI applications, cloud service providers (CSPs) such as AWS, Azure, and Google Cloud are significantly expanding AI infrastructure, ordering large quantities of H100 and the upcoming Blackwell architecture chips. Second, sustained increase in capital expenditure (CapEx) by cloud giants. According to the latest financial reports, the four major cloud providers have increased their AI infrastructure investments by over 50% year-over-year, directly boosting NVIDIA's order visibility. Third, the shift of AI demand from training to inference. While market focus over the past year was primarily on the computing power needed to train ultra-large models, as more models enter the productization phase, inference is accounting for an increasingly larger share of computing demand, creating ongoing contribution to NVIDIA's product lines (especially L40S and L4).

Taken together, this 70% growth is not merely a temporary demand surge but reflects the long-term construction phase of AI infrastructure from zero to one. Unlike past cyclical semiconductor booms, this demand has clearer multi-year characteristics, which is why the market regards NVIDIA's financial results as the "AI earnings barometer"—it reflects not only the company's own performance but also serves as a leading indicator for the entire industry chain's prosperity.

Angle Two: Market Impact – NVIDIA as the "Engine" of the AI Supply Chain

NVIDIA's leadership in the AI chip market makes it a bellwether for upstream suppliers. When NVIDIA's orders grow, direct beneficiaries include: TSMC's advanced packaging (particularly CoWoS technology), high-bandwidth memory (HBM) suppliers like Samsung and SK Hynix, and server contract manufacturers such as Quanta and Wistron. Specifically, each NVIDIA H100 chip requires approximately six HBM3 memory stacks and advanced packaging technology, meaning that for every 10% growth in NVIDIA's revenue, corresponding packaging and memory demand increases synchronously.

This impact extends beyond the hardware layer. NVIDIA's CUDA software ecosystem also fortifies its moat effect. As more developers build applications based on CUDA, switching costs to other platforms increase, further strengthening NVIDIA's pricing power. Therefore, when the market observes NVIDIA's forecast, it is actually assessing the health of the entire AI supply chain—from wafer foundry, packaging and testing, to memory and substrate suppliers, all of which move in tandem with NVIDIA's order growth.

Angle Three: Practical Advice for Readers – How to Judge Whether AI Stocks Are Overheated?

For general investors, simply looking at NVIDIA's 70% growth can easily lead to acrophobia. A more objective approach is to compare "forecast growth rate" with "stock valuation." Taking the PEG ratio (price-to-earnings ratio divided by revenue growth rate) as an example, when the PEG value is significantly below 1, the stock may be undervalued; conversely, it may be overheated. According to the latest data, NVIDIA's price-to-earnings ratio is approximately 60x, with a revenue growth rate of 70%, resulting in a PEG of about 0.86, indicating that even under high growth expectations, its valuation still retains a certain margin of safety.

Nevertheless, investors should still pay attention to the importance of diversification. While the AI industry chain looks promising, a single company or single technology route may face sudden risks. It is recommended to diversify investments across: wafer foundry (TSMC), advanced packaging (Wafer Works), memory (Nanya Technology), and server components (ASUS, Quanta), thereby capturing the upside of the AI boom while reducing portfolio volatility when a single link encounters problems.

Frequently Asked Questions (FAQ)

1. How is NVIDIA's 70% forecast growth calculated? Which businesses does it include?

This growth is calculated by comparing Q2 2026 revenue with Q2 2025 revenue. The main contribution comes from the data center business (accounting for approximately 88% of total revenue), while the remainder comes from gaming, professional visualization, and automotive businesses.

2. Why does the market treat NVIDIA's earnings report as the "AI earnings barometer"?

Because NVIDIA is not only a leading supplier of AI chips, but its order visibility can forward-reflect capital expenditure intentions of cloud giants and enterprise customers. Additionally, the moat effect of its CUDA software ecosystem keeps market share stable, making financial report fluctuations less affected by short-term volatility and more able to reflect long-term industry trends.

3. How significant is the impact of export controls on NVIDIA's subsequent growth?

US export controls on advanced AI chips to China will directly affect NVIDIA's sales in the Chinese market, but considering that China accounts for less than 15% of NVIDIA's data center revenue, and the company already has buffered products for the Chinese market (such as the H20), the actual impact may be partially offset.

4. How will the competitive landscape with AMD evolve?

AMD's share in the data center GPU market is currently still below 10%, although its MI300 series has certain advantages in energy efficiency ratio. However, the CUDA ecosystem's moat effect and long-term software adaptation costs will be the main obstacles for AMD to gain market share.

5. What leading indicators should investors monitor to predict NVIDIA's future performance?

Key indicators to watch: capital expenditure guidance from cloud giants (especially Microsoft and Google), H100 shipment cycle reports, and TSMC's CoWoS capacity utilization rate—these three indicators often lead NVIDIA's revenue by one quarter.