Why Did NeoCloud Experience the Largest Gain During This Round of Tech Stock Rebound in the U.S. Market?

In this round of the US tech stock rebound, one of the strongest directions comes from NeoCloud: CoreWeave, Nebius, and AI infrastructure companies that partially own power and data center resources.

Logically, funds are pricing a multi-leveraged AI infrastructure equity certificate: compute capacity that is already contracted and ready for rapid deployment.

Once AI demand increases, NeoCloud's revenue outlook, financing capabilities, and shareholder equity value could all rise simultaneously. This gives it a strong upward resilience during the tech stock rebound phase; the leverages of power, data center, financing, and valuation together form this layer of leverage.

The bottleneck of AI is changing. Initially, the most scarce resource was GPUs, followed by HBM and high-speed networking; today, what customers truly lack is a complete online-ready capability: access to GPUs, sufficient power, completion of data center construction, network interconnection, and the ability to deliver a scaled cluster within a few months.

NeoCloud is stuck in this gap.

Funds are Buying "Electrified Compute Factories"

Products provided by NeoCloud typically include GPU clusters, networking, liquid cooling, data centers, power access, and operational services. Customers are buying a large-scale compute capacity that can directly run AI training and inference.

This is crucial. While GPUs can be purchased, power capacity, land, substations, data center permits, and network access cannot be replicated in the short term. Large cloud providers have capital and customers, but they are also constrained by construction timelines; some AI companies want to retain more flexibility and are unwilling to put all their demands on a single hyperscaler.

Therefore, NeoCloud, with ready-made power and rapid deployment capabilities, has become an "accelerator" for AI infrastructure investment.

The market is willing to give them a higher valuation, with the core characteristic of such resources being:

· Scarce: Available power and deliverable data center capacity are limited;

· Contractible: Customers are willing to sign long-term contracts with minimum commitment capacity.

When scarce resources can be locked in with long-term contracts, the market will reframe it from ordinary IT service revenue to cash flow assets with infrastructure attributes.

Earnings Report Changed Market's View on Business Model

Prior to this, the market's main question about NeoCloud was straightforward: Buying GPUs and building data centers require massive Capex, will the company fall into a cycle of "continuous financing, continuous burning of cash"?

The recent financial reports have provided a rather positive outlook.

CoreWeave's Q2 revenue reached $2.575 billion, with a disclosed backlog (signed but not yet confirmed expected revenue) of approximately $104 billion; Nebius's AI Cloud ARR (Annual Recurring Revenue) reached $3 billion and disclosed several large long-term contracts. While the market pays attention to quarterly revenue, the focus is more on the complete business cycle behind these numbers:

AI clients signing long-term capacity contracts
→ Some clients providing prepayments or minimum payment commitments
→ The company gaining easier access to debt and equipment financing
→ Additional GPU, data center, and power capacity coming online
→ Revenue growth aligning with EBITDA (Earnings Before Interest, Taxes, Depreciation, and Amortization) growth
→ Improved financing and scaling capabilities

This transformation has shifted NeoCloud's narrative from a "high Capex GPU lessor" to a "AI infrastructure operator expanding based on orders support."

As long as orders, financing, and delivery can sustainably align, growth exhibits a clear flywheel effect.

Why Has Funding Not Primarily Chosen Storage and the Big Three Clouds?

Funding decisions reflect differing expectations in various sectors.

The storage leaders have benefited from AI demand, with strong momentum in products like HBM and DRAM. However, the market has started to worry about supply catching up, high prices, peak profit margins, and whether early optimistic expectations are already fully priced in. Strong financial results can lead to price pressure if the forward guidance is not further revised upward.

The challenge for storage companies lies in their cyclical nature. The market trades on the future trajectory of prices, shipments, and gross margins for the next few quarters; when supply might meet demand, and average selling prices may decline, robust current performance can struggle to sustain valuation expansion. HBM/DRAM, NAND/SSD, HDD also have different sub-cycles, so the stock performance of all storage companies cannot be attributed to a single reason.

The Big Three Clouds—Microsoft Azure, Amazon AWS, Google Cloud—possess more stable cash flows, clients, and technological capabilities and are key beneficiaries of AI investments. Their AI business is diluted by large revenue bases in advertising, enterprise software, e-commerce, consumer businesses, etc.; additional AI capital expenditure also takes longer to translate into overall group margin improvement. For funds seeking resilience, a large NeoCloud contract often has a greater marginal impact on revenue and valuation than an equivalent-sized order would have on the overall valuation of the Big Three Clouds.

NeoCloud is positioned between the two: with a low revenue base, a pure AI exposure, rapid order growth, and the potential for each new long-term contract to directly support the next round of financing and expansion. The market easily views it as a high-resilience AI infrastructure target.

The current market trading logic can be summarized as:

NeoCloud, Essentially an AI Infrastructure Lever

Understanding the leadership of NeoCloud is key to understanding its leverage effect. Buying the stock of these companies essentially means holding an equity asset that is highly sensitive to AI computing power demand, deliverable capacity prices, and the financing environment. The leverage here includes three layers of meaning.

The first is operational leverage. The high upfront investment in GPUs, data centers, power access, network, and maintenance is high, with many costs relatively fixed after capacity comes online. As the utilization rate of the deployed cluster increases and the unit capacity price improves, additional revenue can quickly translate into profit, and the margin improvement in profitability will be significant.

The second is financing leverage. Long-term contracts, non-negotiable commitments, and customer prepayments can enhance the attractiveness of projects to lenders and equipment financiers. The company can thus leverage a portion of its equity capital to leverage larger-scale GPU, data center, and power investments; once the new capacity starts billing, revenue can support the next round of construction.

The third is equity leverage. NeoCloud's revenue base and market value are usually smaller than the big three clouds, but the proportion of fixed assets and debt on the balance sheet is higher. If a large contract simultaneously increases revenue expectations, utilization rates, and financing availability, the market's reassessment of shareholder equity value can be very steep. The rapid upward movement of the stock price after the financial report often comes from a combination of profit forecast upgrades and valuation multiple expansions.

The three layers of leverage constitute a positive feedback loop in the upward phase:

Larger long-term contracts
→ Easier access to financing and capacity expansion
→ Higher utilization and operating profits
→ Increased equity value and financing capability
→ Winning more contracts and opportunities for the next round of expansion

The same mechanism will also amplify downside risks. If customers delay, utilization rates decrease, GPU or power delivery lags, or debt costs rise, fixed costs and financing obligations will squeeze shareholder returns. Therefore, the market's high-resilience pricing of NeoCloud also reflects its high execution requirements.

Order Visibility is the Core of This Round of Revaluation

The most attractive part of NeoCloud is the visibility of its revenue.

If a customer signs a take-or-pay contract, they must fulfill a minimum payment obligation even if their actual usage fluctuates in the short term. For the operator, this type of revenue is easier to predict; for the creditor, these contracts also enhance the feasibility of asset financing.

Therefore, the market will continue to track several indicators:

· Term, enforceability, and customer credit of signed contracts;

· Delta between activated MW (megawatts) and contracted MW;

· Unit MW revenue versus unit MW Capex;

· Proportion of customer prepayments and payment cadence;

· Utilization rate, renewal rate, and customer concentration;

· Debt interest rate, debt term, and subsequent financing ability.

Among these, "activated capacity" is particularly critical. Contracted MW represents demand, and only MW that is energized, installed, and billing will enter revenue and cash flow.

This Is Also a Reassessment of Power Assets

One of the most valuable insights in the NeoCloud community is shifting the focus from GPU quantity to Power (electricity capacity).

GPU supply expands with NVIDIA, AMD, and cloud provider purchases; however, the formation of high-quality power capacity is slower. It involves the grid, substations, land, permits, data center construction, and regional grid conditions.

Whoever can access sufficient power earlier can convert GPUs into sellable computing power sooner.

This is also why some companies transitioning from Bitcoin mining can enter this space: they already have some power resources, land, and infrastructure, only needing to shift assets from mining loads to AI loads. Of course, resource foundation does not guarantee commercial success; ultimately, it still depends on customer, financing, and delivery capabilities.

NeoCloud leads the tech stock rebound, driven by the market's reordering of the AI infrastructure value chain.

The most valued asset at present is the computing power capacity that can combine GPUs, power, data centers, and long-term customer contracts and deliver quickly. It undertakes AI capital expenditure, possesses more contractual features than just chips and components, enjoys high-growth elasticity, and commands a premium for infrastructure scarcity.

Next, whether NeoCloud can continue to outperform depends on a very simple question: can these massive orders be timely transformed into energized clusters, confirmed revenue, and cash flow covering capital costs.