Author: BIT Brokerage

Last night's CPI data was typical—3.4%, in line with market expectations, and did not cause significant ripples at the macro level.

However, there were plenty of highlights in the U.S. stock market last night. The most attention-grabbing focus was on which trading sector could carry the AI narrative after the self-storage chip story temporarily cooled. Aside from optical communications, the earnings reports of several companies last night are indicating to the market that Neocloud (next-generation AI cloud) is becoming a strong candidate.

1. Doubling Revenue and a Tripling Order Book

In the U.S. stock market last night, two spotlight companies in the Neocloud sector, CoreWeave (CRWV) and NEBIUS (NBIS), delivered impressive earnings reports, with both companies ultimately rising by 19% and 34%, respectively.

First, let's look at CoreWeave. Its quarterly revenue reached $2.58 billion, doubling from the same period last year. Even more astonishing is the backlog: it has reached $104.2 billion, expanding two and a half times in twelve months; and the third quarter has just begun, with new contracts adding another $25 billion.

Nebius's growth is even more significant: its quarterly revenue reached $582 million, a year-on-year increase of 454%; its AI cloud business annualized now runs at $3 billion; and in operational terms, EBITDA turned positive for the first time.

When put together, the earnings reports of the two companies point to one fact: existing computing power capacity is completely booked by customers, and there is still a supply shortage.

2. The Wall of This Business Is Not Chips, It's Electricity

It is noteworthy that the way these new cloud providers are building barriers is shifting from "stockpiling chips" to "stockpiling electricity."

The logic is easy to understand: when there is a GPU shortage, placing orders at a higher price can always solve the problem; however, substations, transmission capacity, and grid connection permits cannot be acquired quickly through investment. The chip gap is a business issue, while the electricity gap is a physical issue.

Nebius's recent transaction prices are the best illustration: short-term, rapidly delivered computing power is priced at $40 million to $50 million per megawatt; while long-term contract prices are only $20 million to $25 million per megawatt—the pricing for short-term orders is about twice that of long-term contracts.

The company's sales strategy is even more interesting. For 2027's capacity, customers are now willing to buy entire orders, but the management deliberately holds back some contracts from being signed. It's not that they can't sell, but as electricity becomes more scarce, the retained capacity will be worth more later—the power of pricing should rest in their hands.

3. Burry's Short Formula, Disproven Item by Item by Reality

These two earnings reports also conveniently dismantled a bear case.

According to public reports, short-seller Michael Burry previously questioned whether some manufacturers depreciate GPUs over five to six years, while their actual usable lifespan may only be two to three years, thus suggesting that the profits in this sector might be overstated.

However, the details disclosed by the two companies during the conference call almost entirely overturned this premise.

First, assets have not depreciated. Nebius's first public auction of computing capacity ended with a closing price that was 15% higher than the company's historical highest bid. Something that, according to the bear's argument, "only has two to three years of residual value," buyers are actually bidding up.

Second, the lifespan far exceeds assumptions. CoreWeave confirmed it signed a new contract for A100 computing power at an ideal price, extending the term until 2029. This is a chip that was launched in 2020 and is already three generations behind—according to the bear model, it should have been decommissioned, yet in reality, it has secured a nine-year commercial contract. The company also revealed that the price of older products is not inferior to, or even stronger than several years ago, and the old fleet is essentially fully booked.

Some analysts described the current computing power shortage as: customers are no longer picky; with not enough top-tier new cards available, they are taking all the old models.

4. The Key That Bears Overlooked: Renewal Revenue Equals Net Profit

If the first two points merely overturned the "lifespan assumption," then the third directly breaks through the bear's valuation model.

These old clusters that were renewed have already paid off all asset loans during the initial contract period, and depreciation has also been fully accounted. In other words, every bit of revenue from renewals only faces costs of electricity and minor operational expenses—there is almost nothing else to consume it.

A contract signed until 2029 for A100 equals showing two cards at once: the commercial lifespan of the old chip is much longer than the bears thought; and these assets that were counted as "zero" in the bear model are continuously generating near-pure profit cash flow. Burry not only got the depreciation lifespan wrong but also overlooked the entire second segment of revenue that is the fattest in this business model.

5. Conclusion

Linking the recent market trends, the power transition within the AI sector has become quite clear: storage expectations peaked first, optical communication is taking over with performance trends, and now Neocloud brings forth the triple validations of "sold-out capacity, electricity pricing power, and collapse of the bear logic."

For investors, the upcoming stock selection criteria can actually be condensed into three sentences: Do you have electricity, is the order visibility high, and can old assets continue to generate money? Those that meet these three are solid players in the new narrative; while those that only tell stories and burn capital will be rapidly eliminated in this round of "audit-style" market.

In the second half of AI trading, sheer volume won’t help; you need to have real assets in hand.

Risk Warning: This article is contributed by a third-party author, representing the author's personal opinion and does not reflect BIT's stance or viewpoint. The market data, company earnings information, and analytical judgments involved in this article are sourced from public channels, and BIT does not guarantee their accuracy, completeness, or timeliness, nor does it take responsibility for any losses incurred due to the use of this article's content. The content of this article is for reference only and does not constitute any investment advice, offer or recommendation to buy or sell any financial products. Markets carry risks, investments need caution, past performance does not represent future returns, and investors should independently assess their risk tolerance and take responsibility for their investment decisions.

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。