NVIDIA's $500 Billion AI Financing Plan Hides Concerns: Trump Advisor Warns of 'Dark GPU' Oversupply Risk

NVIDIA's grand plan to bundle GPUs into financial assets is facing risk warnings from within the White House.

Recently, David Sacks, a member of President Trump's Council of Advisors on Science and Technology, stated on the "All In Podcast" that the greatest threat to NVIDIA's plan—collaborating with investment institutions to turn GPUs into financeable, income-generating assets—is not insufficient demand, but an oversupply of computing power. He warned that a "dark GPU" scenario would cause a systemic shock to the entire AI infrastructure investment chain.

Sacks' remarks represent the most direct risk warning regarding this AI financing frenzy from within the Trump administration circle to date.

What is the "Dark GPU" Risk?

Sacks used the "dark fiber" phenomenon from the Dotcom Bubble era as an analogy to explain the scenario he fears.

In the early 2000s, telecommunications companies laid massive amounts of fiber optic cable, only to find demand far below expectations. A large amount of fiber went unused, and prices collapsed, a situation historically known as the "dark fiber" crisis. Sacks believes the GPU market faces similar risks.

He stated directly on the program: "For me, the biggest risk is not on the demand side; the biggest risk is an oversupply of computing power and overconstruction. Just as dark fiber emerged after the internet bubble burst, if 'dark GPUs' appear, it will be a disaster for everyone—especially for those who expected spot prices of $30 to $50 per watt when building computing infrastructure."

The "$30 to $50 per watt" figure refers to Elon Musk's previous public estimate of the value of AI computing power. According to reports, Musk stated during an internal SpaceX conference call that the value of AI computing power is approximately $30 to $50 per watt. Based on this, he projected that providing 1 gigawatt of computing power could generate $300 billion to $500 billion in revenue by the end of 2027.

However, computing infrastructure provider Nebius later disclosed that the annual contract value of its multi-year cloud service agreements is approximately $20 million to $25 million per megawatt. This figure implies that the price Musk anticipated actually corresponds to premium levels for short-term agreements, rather than industry-wide pricing.

Political Resistance as a "Natural Insurance"

Sacks did not stop at issuing warnings; he also offered a counterintuitive judgment: the significant political resistance currently facing data center construction may actually protect the market from oversupply.

He said, "If suddenly too many people rush to supply computing power, leading to oversupply and a market crash, that would be the real risk. But strangely, all this political resistance acts as insurance against that outcome. Because building data centers is extremely difficult—for various reasons we have discussed, there is now a moral panic, hysteria, and even what could be called a farce playing out. It is precisely these political resistances that, I believe, almost guarantee that there will be no oversupply relative to the exponential growth in demand. So, in a strange way, you are actually protected."

In other words, Sacks' logic is: the higher the barriers to construction, the harder it is for supply to expand rapidly, and thus the lower the risk of oversupply.

The Core Logic of NVIDIA's Financing Plan

During the podcast, Sacks also explained the fundamental motivation behind NVIDIA's financing plan: the capital expenditure gap for downstream buyers has become so large that it cannot be covered by equity and debt financing alone.

He cited Musk's expansion plans as an example: "Musk plans to add approximately 6 to 8 gigawatts of computing power next year, and we know this requires $300 billion to $400 billion in capital expenditure. His companies have just raised $100 billion through equity and debt financing. Clearly, they still need to raise funds through other means."

Sacks believes the most direct solution is to apply for seller financing from NVIDIA. He said, "Now, Jensen (Jensen Huang) is creating a credit facility, leveraging major banks and large private equity firms, to make this credit line available to all downstream buyers, which will benefit them."

The core mechanism of NVIDIA's current plan is to transform GPUs into an asset class similar to financial securities, with NVIDIA providing residual value support, thereby making GPUs financeable and income-generating. Sacks characterized this as a key measure to alleviate the current financing bottlenecks in AI infrastructure construction.