AI Capital Frenzy Encounters Market Cold Wave: Computing Power and Clearing Proceed in Parallel

CN
2 hours ago

In the opening scene of this round of the AI capital arms race, on one side, Apollo and Goldman Sachs are pricing a $44 billion GPU financing for SpaceX, betting on NVIDIA's next-generation Vera Rubin chip, planning to launch ground computing power expansion in a leasing model after delivery in the first half of 2027; on the other side, U.S. Treasury Secretary Becerra has identified Anthropic in its IPO as the "national AI champion," emphasizing its compliance with security measures. On November 20, Musk publicly stated, "We will not offer open-source AI models," and coupled with the U.S. Senate passing the DEATH BETS Act, which delineates a regulatory red line for prediction contracts related to war and death, the themes of computing power financing, national team narrative, and contraction of open source are almost simultaneously accelerating along the same timeline. Outside the spotlight of capital and regulation, the MiniMax token dropped 16.6% to 11.95 USDT in the last 24 hours, AI-themed U.S. stocks such as NBIS, AGPU, MRVL, AXTI, and LITE recorded double-digit declines in USDT terms, and Lookonchain noted that approximately 262,976 traders were forcibly liquidated within 24 hours, totaling about $1.54 billion in liquidations. The interplay of super-large computing power financing in the primary market with the price cold wave and concentration liquidation in the secondary market foreshadows that this round of AI and speculation narrative will be forcibly corrected by regulatory red lines and a deleveraging cycle.

$44 Billion GPU Financing Bets on SpaceX Computing Power Leasing

At the same time as passive deleveraging in the secondary market, Apollo and Goldman Sachs have made a proactive leveraged super bet: according to a single source, the two institutions recently announced the completion of pricing for a $44 billion GPU financing for SpaceX, earmarked for its ground computing power construction. This is not an abstract "AI concept financing," but is locked in on NVIDIA's next-generation Vera Rubin chip—Apollo states that this batch of chips will be delivered in the first half of 2027 to build SpaceX's ground computing power cluster, which will later output computing power through a leasing plan. In other words, the most scarce and storytelling asset in today's capital market is no longer the future model valuation of a company but a GPU inventory that has not yet left the factory but has been pre-financialized.

This $44 billion financing, its transaction structure, interest rate, term, and the division of roles among participants have not yet been disclosed. The lack of information somewhat amplifies its symbolic significance: investment banks and large asset management institutions are viewing computing power as a foundational asset that can be packaged, layered, and discounted based on leasing cash flows, while SpaceX assumes the role of "first locking chips, then selling on lease.” The combination of the leasing plan and investment bank structure shifts the stakes in the global AI computing power arms race in the coming years from "who can buy more GPUs" to "who can hold and lease GPUs at a lower capital cost over a longer term." In this logic, the $44 billion pricing is not only an aggressive advance on SpaceX's computing power leasing business but also a collective bet on the future tightness of computing power prices, pulling the AI narrative back from the model and application level to the most primal struggle between electricity and chips, further reinforcing computing power itself as a financial asset.

Anthropic Endorsed as National AI Champion

As computing power is priced as a growth financial asset, the U.S. has begun to pick narrative protagonists who can "support" this asset. Recently, when asked about Anthropic's IPO process, U.S. Treasury Secretary Becerra gave an almost unadulterated evaluation—Anthropic is "certainly" the national AI champion of the United States. This public statement, conveyed through a single medium, essentially writes an up-and-coming laboratory that is still advancing its IPO, with key details yet to be disclosed, directly into the main line of the national industry story, telling the market who represents the U.S. and who holds the priority interpretation at the institutional level.

More importantly, Becerra was not praising model performance but emphasizing that Anthropic has "always cooperated with U.S. security measures." In this discourse, security compliance is no longer just a chapter of risk disclosure but becomes the ticket for advancing to the "national team": those who can actively embed themselves within the U.S. security framework are more qualified to wear the hat of "national AI champion." Therefore, Anthropic's IPO is endowed with dual significance; on one hand, it acts as a starting point for an advanced AI laboratory to seek a higher valuation in the public capital market, and on the other hand, it is a signal released by regulators to the entire industry: the future market will not only reward technological leadership but will also append premiums to companies that are deeply tied to security requirements. This endorsement, while elevating the financing imagination, also raises the regulatory threshold for them to fulfill their security commitments.

Musk Rejects Open Source AI Models, Closes Down Perception

At the same time as the "national AI champion" was officially named, the wind direction on another extreme is quietly reversing. On November 20, Jinse Finance and Odaily Planet Daily reported that Musk publicly stated, "We will not provide open-source AI models." This statement nearly severed the last fantasy of outsiders regarding his potential fluctuation on the issue of open source. Although the specific occasion and complete context of this statement have not been disclosed, making it impossible to determine whether he is talking about all models or a specific part, for developers who have become accustomed to seeking a narrative of technological utopia from Musk, a clear rejection of open source is itself a sufficiently clear dividing line.

For a long time, open-source models have been seen as an important mechanism for increasing transparency, reducing trial and error costs, and accelerating innovation diffusion. The market had hoped that Musk would leave room for open source at least on some level while laying out computing power and models. Now, the signal of "we will not provide open-source AI models," combined with his aggressive expansion on the computing power front, resonates with the trend of computing power being concentrated in a few giants and model companies backed by the national team narrative: the contraction of open source means it is harder for marginal developers to access high-performance models, and ecological innovation is likely to be locked within a few entities that control massive GPUs and compliance endorsements. A highly concentrated AI power structure regarding both computing power and discourse is beginning to take shape.

MiniMax Plummets, U.S. Stocks Fall, $1.54 Billion Liquidated

Just when the primary market was celebrating the $44 billion GPU financing, and policies and the "national team" narrative were continuously ramping up, the secondary market gave a completely opposite response. In the past 24 hours, according to a snapshot from a single source, the MiniMax token's drop expanded to 16.6%, currently reported at 11.95 USDT; several U.S. stocks deeply tied to the AI narrative and priced in USDT declined in sync: NBIS dropped 16.2% to 88.88 USDT, AGPU fell 14.5% to 5.859 USDT, MRVL decreased 14.2% to 127.09 USDT, AXTI was down 13.7% to 30.94 USDT, and LITE also recorded a double-digit drop of 13.6%. In the primary market, computing power financing, regulatory statements, and contraction of open source constitute an increasingly heated narrative chain; meanwhile, these targets are being collectively sold off in trading terminals, with prices directly casting votes of reservation on the prospects of "concentrated power + closed models."

Colder realities come from the collective collapse of leveraged positions. Lookonchain monitoring shows that in the past 24 hours, approximately 262,976 traders were liquidated, involving about $1.54 billion in funds. Although this short-term data derives from a single source, it sketches a scene of high-leverage funds being concentratedly harvested amid an emotional reversal: as the computing power leasing plan points to 2027, leading model companies receive "national AI champion" endorsements, and the open-source window is further narrowed, secondary market funds choose to reduce leverage and cut positions in sync. With real margin and liquidation prices, they are timely correcting the current AI capital structure, indicating that in this competition for concentrated computing power and discourse, those who are really bearing the short-term risks are the traders caught up in it, not the storytellers.

DEATH BETS Act Tightens Death Bets

As high-leverage funds are being concentratedly liquidated in the market cold wave, the U.S. Senate chooses to address the boundary of "betting risks" on the other end. The recently passed DEATH BETS Act (according to a single source from Cointelegraph) directly targets prediction contracts related to war and death, aiming to set an uncrossable regulatory red line for such "death bets." It is not a blanket ban on the entire prediction market but makes clear ethical judgments on the most controversial categories: speculating on prices and election outcomes is permissible, but it does not encourage designing intricate financial games around the end of life and outbreak of wars.

For the crypto prediction market, this means that the niche categories originally attracting traffic via "extreme event topics" are being excluded from options by the system. The bill is seen as a response and risk control measure from regulatory authorities regarding death bet contracts but has not yet publicly disclosed specific terms, effective dates, and transitional arrangements for existing contracts. However, the trend is clear: in an era where AI-driven data and prediction models are overwhelming, and any event can be quantified into probabilities and odds, the system is beginning to actively delineate for extreme speculation, distinguishing between what "can be calculated" and what "can be bet on.” This bill serves as a reminder for all those betting on the future that while technology can expand infinitely, society's tolerance for speculation on death and war ultimately has boundaries.

Primary Hot, Secondary Cold: The Next Act of the AI Arms Race

From SpaceX's $44 billion Vera Rubin computing power financing, to Anthropic being dubbed the "national AI champion" by the Treasury Secretary, to Musk making it clear on November 20 that "we will not provide open-source AI models," power and capital are converging towards a few computing power platforms and model companies. The AI arms race is heating up in the primary market with national team endorsements, closed technology stacks, and massive chip orders. At the same time, the MiniMax token plummeted 16.6% to 11.95 USDT, while U.S. stocks like NBIS, AGPU, MRVL, AXTI, and LITE recorded double-digit declines, with around 262,976 traders swept out in $1.54 billion liquidations. The secondary market rapidly cooled amid severe corrections and forced liquidations, laying bare the structural mismatch of "primary hot, secondary cold." As the DEATH BETS Act delineates hard boundaries for extreme speculation, the financialization of computing power, the national team model consolidation, and the speculation red line will reshape the migration route for funds and developers over a longer period, making the next chapter of AI both a centralized arms race and a collective redefinition of risk tolerance.

Join our community to discuss together and become stronger!
AiCoin exclusive Hyperliquid benefits: https://app.hyperliquid.xyz/join/AICOIN88
AiCoin exclusive Aster benefits: https://www.asterdex.com/zh-CN/referral/9C50e2
On-chain Telegram community: https://t.me/AiCoinWhaleData
On-chain community: https://www.aicoin.com/link/chat?cid=N6OVMor5g
AiCoin on-chain Twitter: https://x.com/aicoinwhaledata

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[email protected],本平台相关工作人员将会进行核查。

Share To
APP

X

Telegram

Facebook

Reddit

CopyLink