One morning, Meta's Zuckerberg did something significant: he announced that Meta would officially sell its AI computing cloud service to the public.
As soon as the news broke, the chip sector in the U.S. stock market faced a fierce sell-off:
Philadelphia Semiconductor Index: plummeted 6.27% in a single day
Micron: collapsed 10.57%
SanDisk: collapsed 10.62%
Intel: collapsed 9.03%
Corning: collapsed 13%
A computing storm initiated by Meta is drawing the entire AI hardware industry into a heated debate about the "demand ceiling".
1. Meta's "Turnaround": From Burning Cash to Selling Water
To understand why the market reacted so strongly, we need to look at what Meta has been doing in recent years.
Over the past two years, Meta has been one of the most aggressive "buyers" in the AI arms race. Spending $125 billion to $145 billion annually on capital expenditures, it has been frantically purchasing GPUs, networking equipment, optical modules, power, and cooling facilities—solely to catch up with OpenAI and Anthropic on the large model track.
The problem is, the money has been spent, but the models have not been developed. Meta's own large models have consistently lagged behind those of major competitors, resulting in a significant amount of deployed computing power being idle or operating inefficiently.
To illustrate, this is like a marathon in the AI era. Meta, initially racing hard on the track, discovered over time that it was drifting further away from the leading competitors, so it sat down at the roadside and began selling water to passersby.
At that moment when it sat down, it became a "computing power seller".
Specific data can better illustrate how rich it is:
By the end of 2025, Meta's AI computing power will be equivalent to approximately 2.5 million H100s, totaling approximately 2GW of power scale
In 2026, Meta's projected capital expenditure guidance is $135 billion, corresponding to an additional approximately 2-3GW of computing power
Simple calculations suggest that by the end of 2026, Meta's total computing power could approach 5GW
What does this signify? It means that the AI computing capability Meta possesses may have already surpassed that of any technology company besides ultra-large scale cloud providers. Moreover, it cannot even use all of it.
Therefore, the logic of renting out idle computing power makes sense—it is wasteful to have assets sitting unused, monetizing them at least makes the accounts look better.
2. Why Are Chip Stocks Falling? Two Market Voices
Regarding Meta's selling of computing power, there are currently two different interpretations in the market.
Interpretation One: Bearish—The "Excess Computing Power" Signal Has Been Fired
The bearish logic is straightforward:
Meta has been one of the largest buyers of AI chips over the past two years. Now it is starting to rent out excess computing power—this indicates it no longer needs to buy more.
If even a player like Meta realizes they "overbought," what about other manufacturers? Is the large model arms race transitioning from "full sprint" to "slamming the brakes"?
A further deduction is that once the consensus of "excess computing power" forms, the entire AI hardware market's demand growth expectations will need to be significantly downgraded. Core players in the supply chain such as Samsung, TSMC, Micron, and NVIDIA may face risks of slowing order growth or even order cuts. The hardware industry story that has been thriving for over two years might begin to discount.
Interpretation Two: Bullish—"Selling Water" is to Continue Running the Marathon
The bullish counterargument is equally compelling:
The hundreds of billions spent by Meta on GPUs over the past two years are sunk costs that have already been incurred. Now, turning these idle assets into cash is not abandoning competition but a return to business rationality.
If renting computing power can truly generate profit, then Meta's subsequent purchases of GPUs, networking equipment, and optical modules could be even more robust—because having the ability to recover funds allows for more sustainable spending. Spending money on devices → renting out idle computing power to recover funds → using the recovery money to buy more devices, this is actually a positive cycle, rather than a zero-sum game.
3. Who are the Winners? Capital Offers a Clear Answer
The two viewpoints are currently in fierce contention, and short-term fluctuations in chip stocks are likely to persist. But one thing is certain—the capital market is rewarding Meta for the decision to "sell water" with real money.
On the trading day following the announcement, Meta's own stock price did not plummet along with the chip stocks. Instead, it surged nearly 9%. This indicates a clear attitude from investors: regardless of how chip stocks fall, this move by Meta is a positive for itself.
Why is this the case?
The core issue is not "how much money can be made from renting out computing power." Because even if Meta rents out all surplus computing power, it's hard to say how much incremental net profit it will bring in the short term. Perhaps the initial figures are only in the range of $2 billion to $3 billion; for a company with annual revenue exceeding $100 billion, this number itself is not critical.
What the market is truly concerned about is the change in attitude.
In the past few years, Wall Street's biggest anxiety regarding Meta has been its "unlimited burning of cash." Annual capital expenditures of $125 billion to $145 billion represent a hole that can never be filled, raising investors' doubts about return on investment (ROI).
Against this backdrop, Zuckerberg's willingness to monetize surplus computing power—even if the amount is not large—conveys a message: management is starting to care about capital efficiency, and the phase of "madly burning money" might be reaching a peak.
This is a signal Wall Street has been waiting for. Therefore, even if chip stocks collapse, Meta's own stock price rose against the trend. The capital rewards are not for the computing power rental business itself, but for the strategic shift from "unlimited arms race" to "return to business rationality."
4. The Real "Black Swan" May Not Have Appeared Yet
Although chip stocks have fallen sharply, it may still be too early to say that the AI hardware bull market has ended.
Currently, the biggest uncertainty, aside from Meta renting out computing power itself, lies in whether other giants will follow suit.
Meta is the first prominent technology company to publicly declare, "We have overbought computing power and need to rent it out." So what about Microsoft and Amazon? If these cloud giants, which are also heavily purchasing AI infrastructure, cannot withstand the pressure from the capital market and announce plans to "invest rationally" and cut capital expenditures—then that will be the moment that truly shakes the foundation of the entire AI hardware industry chain.
So far, Microsoft and Amazon have not released similar signals. But if in the coming weeks or months, "cutting capex" becomes a collective action among tech giants, the value of AI chip hardware will be re-evaluated.
In other words, the current market is shifting from one question to another: it's not about the problem lying with Meta, but whether Meta will be the first domino to fall.
5. In Conclusion: Volatility is Certain, Direction is Uncertain
For investors focused on chip stocks, the current situation can be summarized in one sentence: extremely high uncertainty, short-term volatility will intensify.
How significant will Meta's renting out computing power impact chip demand? Is it an isolated event or the beginning of an industry trend? Will Microsoft and Amazon follow? These questions remain unanswered. Until the answers surface, stock prices in the chip sector are likely to continue fluctuating.
For those wishing to participate in the volatility trading of chip stocks, BIT (formerly Matrixport) platform offers a real trading channel for Nasdaq-listed chip stocks—covering core targets such as Micron, Intel, AMD, and NVIDIA. What you are buying are the genuine shares registered within the DTC system, not tokenized certificates, allowing for trading at any time and transfers via ACATS.
In conjunction with BIT's 24/7 near-instant deposit channel for USDT, even if you are in the Asian time zone, you can complete your funding before the US stock market opens, ensuring you do not miss any important market windows.
The story of AI hardware is not over, but the script is being rewritten.
Disclaimer: This article is for informational reference only and does not constitute investment advice. The market interpretations and data in the text are based on publicly available information and may have delays or inaccuracies. The volatility risks of US stocks and chip sectors are high, and past performance does not guarantee future returns. The specific terms of related trading services are subject to official explanations from the BIT platform, and users in different regions must confirm compliance on their own. Investment carries risks, please consult professional advisors before making decisions, and assume the corresponding risks and consequences.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
