NEWS

In the early hours of March 19th, Beijing time, NVIDIA unveiled a series of AI chip and technological breakthroughs at the 2025 GTC Conference. During the event, NVIDIA revealed that it is ramping up production of the Blackwell chip and will transition to the Blackwell Ultra in the second half of this year. The next annual chip release, Vera Rubin, will begin shipping in the second half of 2026, succeeding the Blackwell Ultra chip. In terms of specifications, these chips have upgrades in process technology, memory technology, power consumption, and core innovations. For example, Vera Rubin features TSMC's 3nm + CoWoS-L packaging and HBM4 (384GB) memory technology, with lower power consumption per unit of computing power. It also introduces CPU-GPU integration and CPO technology, targeting the next-generation quantum-AI collaborative computing market.

However, despite the numerous technological highlights and CEO Jensen Huang's passionate two-hour speech to outline NVIDIA's future blueprint, the capital market did not respond enthusiastically. On the day of the conference, NVIDIA's stock closed down 3.4% and fell another 0.55% in after-hours trading. According to comprehensive market analysis, the main reasons are as follows:

Analysts have also expressed different views on this matter. Maribel Lopez, founder of Lopez Research, stated that quantum computing and robotics, mentioned by Jensen Huang, have already been thoroughly discussed and cannot become NVIDIA's main sales drivers in the short term, nor can their development speed compare with that of the core chip business. Blayne Curtis, an analyst at Jeffreies, hopes to see more evidence proving that NVIDIA has the potential to expand the total market size and provide advantages in terms of cost.

However, some Wall Street figures continue to be bullish on NVIDIA. Will Stein, an analyst at Truist, believes that despite market concerns about NVIDIA's customers entering a computing power digestion period, he still regards NVIDIA as a key player in artificial intelligence. Its leadership position benefits from an innovative culture, ecosystem, and continuous investment in software, training models, and services. Dan Ives, an analyst at Wedbush, stated that the market needs stable policies, and market fluctuations in the coming months will not change the trajectory of the AI revolution. He believes that the GTC Conference is a turning point for tech stocks.

It is worth noting that China's AI industry is showing strong development momentum. With the launch of DeepSeek-R1, domestic AI has reached the same level as the United States in the model field, and there is hope for overtaking on a curve. China's AI industry chain is beginning to show international competitiveness with joint efforts from the policy, technology, and demand sides.

The stock price fluctuations following NVIDIA's GTC Conference not only reflect the market's complex attitude towards its technological development and business prospects but also highlight the intense competition in the global AI industry. The rise of China's AI industry is injecting new vitality into the global AI market, and the future market landscape is worth looking forward to.