As Rubin enters mass production, NVIDIA Corporation (NVDA.US) tightens its 50% revenue-sharing model and addresses market anxieties through "de-circulation" governance.

Recent reports from local media on Thursday, citing informed sources, indicate that NVIDIA Corporation (NVDA.US), which has just released a stunning earnings report, has paused some transactions in its new financing plan. Originally, this plan was designed to offer credit services to cloud computing companies focused on AI computing power infrastructure in exchange for revenue sharing. Informed sources stated that the chip giant temporarily scaled back this plan last week and added that NVIDIA Corporation may redesign it in the future or integrate it into another initiative. NVIDIA Corporation has only halted certain revenue-sharing transactions within the "AI Compute Partnership" and has not terminated the entire program, nor does it imply that demand for NVIDIA Corporation's AI GPU computing clusters, such as Vera Rubin, is weakening. NVIDIA Corporation has clarified that this business model designed to expand the availability of computing power still exists and will continue to adjust due to strong demand. What has been genuinely halted is a part of the most controversial transactionswhere NVIDIA Corporation sells GPU computing clusters while also providing backstop financing for unsold computing power, limiting rental clients, and extracting a portion of the revenue beyond a threshold from over 50% revenue sharing. In the second quarter of FY 2027, NVIDIA Corporation's revenue skyrocketed by 106% year-on-year to $96.2 billion, with data center revenues soaring by 117% to $89 billion, and revenue growth is expected to be around 70% in the next fiscal year. Meanwhile, Anthropic has signed a six-year, $45 billion computing power procurement agreement with Nscale to deploy the latest Vera Rubin platform. These figures indicate that the global demand for AI infrastructure and next-generation computing clusters remains exceptionally strong, and NVIDIA Corporation's pause on certain transactions does not reflect a cooling of computing power orders; rather, it is a risk contraction on its credit support, computing power leasing, and revenue sharing model, actively promoting a "de-cyclization" of its most controversial computing transactions. Financing boundaries are tightening! NVIDIA Corporation has paused some revenue-sharing transactions, leading to a reassessment of its $500 billion credit expansion. More specifically, NVIDIA Corporation is setting up a firewall against the controversies of "circular financing," antitrust scrutiny, and excessive control over customers' operations: while this plan can promote small and medium-sized AI cloud service providers' purchases of NVIDIA Corporation chips, it may also position hardware suppliers as financing supporters, guaranteed lessees, and revenue participants simultaneously. As tech leaders such as Jensen Huang, Sam Altman, and Elon Musk are set to appear together at the G20 technology summit, this adjustment indicates that the super construction cycle of AI computing power is transitioning from merely competing on capital investment to a new phase that values financing quality, regulatory boundaries, and genuine terminal demand. NVIDIA Corporation's suspension of certain transactions that encompass a combination of "selling GPUs + credit support + unsold computing power leasing guarantee + overage cloud revenue 50% sharing" signals a push towards "de-cyclization" of the computing transactions, effectively reducing the closed-loop nature of "financing customers to purchase its own chips and subsequently guarantee revenue," allowing orders to be more reflective of independent terminal demand and actual cash flow, while also avoiding antitrust risks and excessive control over customers operations, as well as risks relating to credit ratings from the "AI circular financing" concern. A spokesperson from NVIDIA Corporation responded to media inquiries regarding the latest reports, stating: "This new business model that opens channels for computing power use in the rapidly growing AI ecosystem... is still operational and continues to evolve due to strong demand for AI computing power." According to media reports, NVIDIA Corporation took this action less than two months after announcing the plan. The initiative aims to meet the strong financing needs of small AI cloud computing firms, as well as the demand for AI computing resources. Previously, NVIDIA Corporation planned that if cloud computing clients could not rent out their full computing capacity, the company would reverse lease those resources, providing a safety net buyer for clients and making it easier for these startup AI cloud computing enterprises to borrow the large sums needed to purchase NVIDIA Corporation's AI GPU clusters. In this model, NVIDIA Corporation would first obtain sales data through hardware sales and would then also share a portion of the cloud computing revenue derived from clients utilizing NVIDIA Corporation's computing power. During the earnings call this week, NVIDIA Corporation indicated that, from a medium to long-term perspective, this sound business model is expected to generate billions of dollars in revenue. However, in recent months, as NVIDIA Corporation continually reinvests funds into the AI ecosystem, investor scrutiny and cautious stances have intensified, raising serious market concerns regarding its involvement in so-called "AI circular financing and transactions." There is worry that NVIDIA Corporation, as a leading force in AI GPUs, may artificially inflate demand, which has led to increased market discussions about a potential "burst of the AI credit bubble," resulting in a short-term sell-off for NVIDIA Corporation and other leaders in the AI computing capacity industry in both stock prices and credit bond trading prices. This month, the company (NVIDIA Corporation) assisted its clients in securing $500 billion in financing from major U.S. Financial Institutions, Inc. and agreed to provide guarantees of up to $105 billion to help OpenAI lease a large data center. Media sources cited informed insiders reporting that some senior employees at NVIDIA Corporation had expressed concerns to existing and potential clients about the possibility of antitrust scrutiny arising from the plan; cloud computing clients have also shown sensitivity regarding the extent to which NVIDIA Corporation can influence their operations, especially among large-scale cloud computing giants. Reports indicated that in the first few weeks of launching the program, the level of control NVIDIA Corporation attempted to impose had already angered some potential major partners. The reports added that NVIDIA Corporation had informed some cloud service providers that they could only lease NVIDIA Corporation's AI GPUs to clients approved by the company; NVIDIA Corporation also expressed a preference for these computing resources to be allocated to multiple smaller businesses or to AI cloud computing leasing firms in their initial stages, rather than concentrating the power with a massive cloud computing client. Informed sources indicated that according to some transaction agreements in the proposed stage, when cloud service providers achieve revenue data from utilizing NVIDIA Corporation's AI GPUs that exceed a certain operational threshold in the agreements, NVIDIA Corporation would acquire 50% of the excess revenue; however, these initial agreements raised doubts among some large cloud computing companies. The AI super bull market has not fizzled out; the financing flywheel is first making subtractions! NVIDIA Corporation's AI GPU computing capacity chain is stepping into the era of "de-cyclization." NVIDIA Corporation's latest move appears to resemble a regulatory and credit risk "proactive brake": temporarily reducing potential revenue sharing, yet helping to ease concerns surrounding circular financing, antitrust scrutiny, and client operational autonomy disputes, enhancing the verifiability of future income, and overall may belong to a neutral slight improvement in governance nature concerning NVIDIA Corporation's robust fundamental outlook. The demand side has not shown a turning point; rather, it continues to exhibit exponential expansion. In the second quarter of FY 2027, NVIDIA Corporation's revenue grew by 106% year-on-year to $96.2 billion, while data center revenues exploded by 117% to $89 billion, with third-quarter revenue guidance reaching $108 billion, plus or minus 2%. Demand for Vera Rubin remains strong and has fully commenced mass production; management projects growth of about 70% in FY 2028 and emphasizes that this figure is constrained by memory/storage chip supply and advanced process manufacturing capacities at Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR. Simon Leopold, a senior analyst at the well-known Wall Street investment firm Raymond James, significantly raised NVIDIA Corporation's target stock price from $352 to $550 in a report on Thursday. Leopold's analysis indicates this target price presents approximately 130% upside potential compared to current levels. Based on this estimate, NVIDIA Corporation's market capital could reach $13 trillion. Currently, this chip giant has a market cap of approximately $5.2 trillion. Meanwhile, Anthropic has signed a six-year, $45 billion computing power procurement agreement, securing about 460 megawatts of power capacity and plans to deploy NVIDIA Corporation's Vera Rubin platform. These signals indicate that the global race for AI infrastructure is extending from chip orders to bottlenecks in core data center infrastructure, power resources, and long-term computing power contracts, with the intensity of capital investment still accelerating. The market's latest description of NVIDIA Corporation as the "AI central bank" or "Balance-Sheet-as-a-Service" accurately captures NVIDIA Corporation's trend of expanding from a chip supplier to an ecological financing intermediary; however, it is not accurate to fully equate NVIDIA Corporation's off-balance-sheet commitments and guarantees totaling $530.5 billion as already formed debt. Official documents from NVIDIA Corporation show that this figure is made up of $366 billion in future commitments, $56 billion in additional AI cloud and third-party data center commitments, and a maximum nominal guarantee exposure of $108.5 billion; among these, $279 billion concerns supply commitments related to memory and manufacturing capacity, some of which can be canceled, postponed, or adjusted, while the $105 billion related to OpenAI's guarantees will be gradually activated starting in FY 2029 across nine construction phases. The real concern lies with cash conversion: accounts receivable rose from $38.466 billion to $63.059 billion, with five major clients accounting for 70% of this; operating cash flow in the second quarter was approximately $24.077 billion, significantly lower than $59.688 billion in net profit under U.S. GAAP. However, the gap is still influenced by equity investment income, inventory increases, and extended payment terms for investment-grade clients and cannot directly equate to worsening client credit. For the entire AI computing capacity industry chain, NVIDIA Corporation's adjustments regarding revenue sharing agreements and computing power credit guarantees are not a turning point in the economic climate but may represent a watershed in financing discipline that alleviates tensions around AI circular financing and antitrust scrutiny. The capacity chain led by Taiwan Semiconductor Manufacturing Co., Ltd. Sponsored ADR, alongside high-bandwidth memory dominated by HBM and high-performance NAND, advanced packaging systems, high-speed Ethernet infrastructure, optical interconnects, data center power chains, and liquid cooling driven by colossal-scale cloud vendors and high-credit clients still have high certainty in their orders; the real pressure will come from small and medium-sized emerging cloud service providers relying on NVIDIA Corporation's credit endorsement, lacking long-term purchasing contracts and stable cash flow, and from high-leverage financing chains formed around speculative data center parks.