Preface: When Computing Power is No Longer the Sole Metric

In 2026, the evaluation system of the Bitcoin mining industry is undergoing a dramatic change. The simple "scale of computing power" is no longer the only standard for determining market value. The capital market is looking for two types of targets: one is giants like Core Scientific that have successfully transformed into AI; the other is miners like CleanSpark that pursue efficiency to the extreme.

Beyond these obvious giants, CANG represents a third, atypical survival model. This article will combine the latest release of Digital Mining Solutions' "2025 Bitcoin Mining Market Review" (hereinafter referred to as the "2025 Annual Report") to conduct a deep horizontal comparison of three representative companies—MARA, CLSK, CORZ—with CANG, dissecting the differences in their underlying business logic and the significant pricing discrepancies.

I. Asset Expansion Model: Heavy Sword Without Edge vs. Guerrilla Tactics

In terms of computing power acquisition costs (CAPEX), the industry has diverged into two distinctly different paths.

1. Heavy Asset Faction: CleanSpark & Riot Platforms

● Core Strategy: "Buy new, not old; build rather than lease." CLSK and Riot tend to invest heavily in building large-scale mining farms and purchasing the latest S21 or XP series mining machines.

● Advantages: Extremely high energy efficiency (low J/TH), long-term operational stability, and strong appeal to institutional investors.

● Disadvantages: Extremely high capital expenditure (CAPEX). The cost of new machines typically ranges from $15 to $25/TH, which extends the return on investment (ROI) period. Once the coin price stagnates, the enormous depreciation pressure can erode profits.

2. Value Arbitrage Faction: CANG

● Core Strategy: "Supply chain leverage, extremely low-cost expansion."

○ Low-cost Positioning: CANG does not blindly chase after overpriced new machines. The 2025 Annual Report specifically points out that CANG rapidly entered the industry's top tier by acquiring second-hand mining machines, obtaining computing power at an extremely low cost (about $8/TH), creating a natural safety cushion compared to peers.

○ Dynamic Upgrades: CANG does not always use old machines but adopts a "Refresh" strategy, for example, upgrading 6 EH/s of computing power to S21 in Q4, only paying the price difference.

● Comparative Conclusion: CLSK excels in efficiency, while CANG excels in the cost-effectiveness of computing power. In the current highly competitive environment, CANG's unique "light asset" model provides greater resilience against risks.

II. AI Transformation Path: International Freight Hub vs. Same-City Express Network

As mining rewards halve, AI has become a battleground for mining companies. Data from the 2025 Annual Report shows that companies with clear AI/HPC revenue significantly outperform pure mining companies in stock price. However, in this arena, CANG and the giants have chosen entirely different entry points.

1. Core Scientific: International Freight Hub

● Business Model: CORZ has signed a massive contract with CoreWeave, committed to building Tier 3/4 level super-large data centers.

This is akin to constructing a vast "international airport cargo terminal." It specifically serves "giant containers" (large model training tasks) with astonishing throughput. However, this type of infrastructure has extremely high requirements for runways and control towers (dual power supply, ultra-high redundancy), with construction cycles lasting several years and a funding threshold reaching billions; once built, it becomes an irreplaceable core hub.

● Valuation Logic: The market views it as "core infrastructure," granting it a very high premium because it possesses irreplaceable scarce resources.

2. CANG: Same-City Express Network

● Business Model: CANG avoids the red ocean of building "airports" and chooses to focus on the long tail of "AI inference." Through low-cost, ultra-fast AI transformation, it converts scattered global mining farms into decentralized AI nodes.

This is like creating a dense "same-city front warehouse" or "express delivery network."

○ Flexible Transformation: There is no need to build expensive airport runways; instead, existing community points (mining farms) are utilized for simple power and network upgrades, achieving plug-and-play.

○ Handling Small Packages: It does not transport giant containers but specializes in high-frequency, rapid processing of massive "small packages" (inference requests from small and medium-sized enterprises, real-time response tasks).

○ Intelligent Scheduling: Through a middleware platform, it schedules scattered global computing power like dispatching couriers, forming a distributed network.

● Comparative Conclusion: CORZ is engaged in "heavy infrastructure," earning expensive tolls but is too heavy and slow; CANG is involved in "last-mile logistics," profiting from high-frequency turnover and flexibility. On the eve of an explosion in AI inference demand, CANG's "express delivery" model, which can quickly respond to the needs of small and medium-sized enterprises, has seized the advantage of "speed" and "breadth."

III. Market Perception and Chip Structure: Crowded Trading vs. Cognitive Wasteland

If the fundamentals determine the long-term floor, then the deviation in market perception determines the short-term explosive power. The institutional holding data disclosed in the 2025 Annual Report reveals the most astonishing contrast.

1. Wall Street Darlings: MARA & IREN

● Current Status: "Crowded Trading."

○ Data: MARA has 484 institutional holders, while IREN has 433.

○ Logic: Almost all funds focused on Crypto or AI have already allocated to these leaders. Their information is completely transparent, and expectations have been fully priced into the stock price. To push the stock price further up, extremely impressive performance surprises are needed.

2. Forgotten Corner: CANG

● Current Status: "Cognitive Wasteland."

○ Data: Despite its computing power ranking among the top five globally (50 EH/s), CANG has only 35 institutional holders.

○ Logic: This means that the vast majority of institutional funds have yet to cover this company. This extreme information asymmetry is often a source of excess returns.

○ Catalyst: As CANG's AI business begins to disclose information, or its Bitcoin holdings are re-evaluated, even a small amount of institutional funds starting to "fill the gap" could bring significant marginal pricing impact for such a low liquidity, undervalued target.

○ Comparative Conclusion: Investing in MARA is following the consensus, earning industry Beta; investing in CANG is anticipating cognitive repair, betting on the Alpha brought by institutional entry.

IV. Conclusion: Investor's Choice

Finally, we return to the core valuation logic.

In-depth Commentary:

The market currently assigns CORZ and MARA a pricing that includes a very high "expected premium." Meanwhile, CANG's market value is even lower than its book value of "hard assets" (BTC holdings + mining machines + cash - debt).

This means that the market has not assigned any positive valuation to CANG's 50 EH/s computing power and Tier 2 AI transformation.

Investment Recommendations:

● If you believe in the effect of capital clustering and pursue absolute liquidity and industry beta, MARA remains the top choice.

● If you are optimistic about the certainty of the AI training end explosion and do not mind high valuations, CORZ is the purest target.

● However, if you are a value investor looking for a high-odds target with asset backing below and AI transformation options above, then CANG clearly offers a highly attractive entry position under the current pricing system.

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。