To The Moon

Once again, we are right about the misinformation and manipulation related to DeepSeek. We have anticipated and corrected that DeepSeek isn’t a threat and was full of misinformation. Here we also seek to provide clarity to the recent misinformation:

JD Vance’s comments

JD Vance, despite being the Vice President, does not have the authority to govern AI decisions at the White House. 

JD Vance stated that the U.S. will safeguard American artificial intelligence technologies from “theft and misuse” by adversaries. However, the perspective driving Intel’s recent surge is actually directed at OpenAI, not Intel, particularly in response to allegations that Deepseek was trained using OpenAI’s ChatGPT. Huawei is already ahead of Intel in chip technology and does not need to steal from them

Additionally, we believe that other tech giants in the AI race will not use Intel chips solely because they are made in the USA. Trump faces a crucial decision: whether to maintain AI leadership or allow the country to fall behind just so Intel can catch up, especially considering that Huawei’s chips are more advanced than Intel’s.

AI leaders like Sam Altman and Elon Musk would not sit back and risk losing the AI lead because they can only use Intel chips. Furthermore, it is important to note that Masa (SoftBank), who backs the Stargate project, is a strong supporter and close ally of NVIDIA. Investors should view this Stargate initiative as a US-Japan collaboration rather than solely a US-Taiwan effort.

OpenAI and Meta self-design chips

OpenAI and Meta are both pursuing custom chip development, with Meta already working with Broadcom while also acquiring a Korean chip startup. 

Each company has its own approach to AGI that necessitates experimentation with custom chips. However, this does not impact NVIDIA, despite the widespread misinformation suggesting otherwise. In reality, this signals increased spending rather than a reduction in investment.

White House’s Perspective

The White House’s perspective is unlikely to change the short-term impact on NVIDIA and TSMC. A relevant example is Foxconn’s factory during Trump’s first term—despite efforts, Apple products are still primarily made in China, with some expansion into India over the years.

If we apply this to AI, JD Vance’s perspective does not align with the reality of global supply chains. The administration faces a delicate balance; if they focus solely on Intel’s success, they risk losing AI leadership. 

Their approach appears to be a long-term strategy that will take years to materialize rather than an immediate shift. This is not a new development, as TSMC was already building factories in the U.S. and will continue to do so, though some setups will take time. Overall, this is more of a continuation of previous efforts rather than a disruption.

NVIDIA’s realities

The market reaction is irrational—there is no scenario where hyperscalers will suddenly start ordering Intel chips while boycotting NVIDIA. Likewise, NVIDIA is not at risk of losing its lead to Intel, especially considering that NVIDIA itself is a US company. 

If there is a need and potential for Intel to become a major US-based foundry, it is more likely that NVIDIA would become one of Intel’s key US clients rather than a direct competitor.