If you don't understand NVDA, you can search for Jensen Huang's background and speeches. Generally, when a founder is still with the company and has personally scaled it to such a size, you should trust his and the company's capabilities.

Even without considering these factors, historically, every time NVDA's price surges, it tends to plateau for a long time. The same happened before the April rally. Before the surge, NVDA dropped from 150 to 90. Starting with DeepSeek, negative news kept piling up, each more absurd than the last. First, it was claimed that DeepSeek's training code bypassed CUDA (it actually used CUDA's low-level language, akin to assembly). Then, it was said that Huawei's chip performance had caught up to NVDA's GPUs (no comment—just refer to the news about DeepSeek V2's training failure six months later). Finally, when everyone was thoroughly demoralized, the stock took off.

This pullback is no different—semiconductor news keeps getting wilder. First, it was the "AI bubble" (anyone still using GPT or Gemini knows how much productivity has been unlocked). Then, it was the "H200 sales ban" (who was it that opposed the ban earlier?). Next came claims that TPUs outperform GPUs, and now it's about lithography machines...

Support technological innovation, but also recognize the objective cycles of R&D and the allure of sensationalism.$NVIDIA(NVDA.US)