"AI Version of Financial Innovation": NVIDIA Corporation's "$500 Billion Grand Calculation"

NVIDIA Corporation (NVDA.US) is reshaping its commercial boundaries in an unprecedented mannertransitioning from a chip manufacturer to an "financial partner" in AI infrastructure, centered around a new cloud service revenue-sharing model based on a "bottom-line guarantee for profit-sharing." NVIDIA Corporation is actively advancing revenue-sharing agreements with emerging AI cloud service providers (neocloud): NVIDIA grants credit endorsement for the initial construction of these cloud service providers in exchange for a percentage of their cloud revenue. According to Windward Trading Desk reports, Morgan Stanley analyst Joseph Moore likened this mechanism to "financial innovation in the AI field," believing its potential scale is considerableshould the neocloud ecosystem develop to a scale of 25 gigawatts of computing power comparable to major hyperscale cloud providers, and with a rental price of $20 million per megawatt, this could generate a revenue pool of $500 billion annually. If NVIDIA Corporation captures a quarter of the profit-sharing, with a gross margin of 100%, it could bring a significant 60% upside to its FY28 earnings before interest and taxes (EBIT). Morgan Stanley maintains an "Overweight" rating on NVIDIA Corporation, ranking it as a preferred pick in the semiconductor industry, with a target price of $288 against the current stock price of $225.30. Analysts point out that this new business model combines high upside potential with limited downside risk, although they also acknowledge that market skepticism surrounding its cyclical revenue structure may further exacerbate the divergence between bulls and bears. Profit-sharing through credit endorsement creates annuity-like income. According to the Morgan Stanley report, NVIDIA Corporation's new neocloud cooperation model operates around a core mechanism: NVIDIA provides credit endorsement to AI cloud service providers by committing to pay a specific hourly rental price for GPUs, enabling them to finance data center construction at near-investment-grade costs. In return, NVIDIA receives revenue sharing from the portion exceeding the guaranteed base price. Morgan Stanley, citing insights from a roadshow with NVIDIA Corporation's CEO, reported that both parties negotiate the guaranteed GPU hourly rental price, allowing AI cloud providers to secure financing based on that guarantee; NVIDIA Corporation earns revenue sharing on any amount rented above the guaranteed price, with specific ratios varying by agreement. The report further notes that in some cases, NVIDIA's endorsement is not purely a cash-backed guarantee but instead involves obtaining cloud providers' GPU usage rights for its own R&Dsimilar to previous arrangements with CoreWeave, but with greater upside potential for NVIDIA. In sensitivity analysis, Morgan Stanley assumes that NVIDIA Corporation receives 35% of the revenue exceeding the guaranteed price. Based on a GB300 data center unit of 1 gigawatt, under scenarios where GPU rental prices range from $10 to $16 per hour, every 5 gigawatts of computing power could generate an incremental annual income of $37 billion to $65 billion for NVIDIA Corporation, boosting EPS by approximately 7% to 13% for FY29. Even at a more conservative scale of 2 to 5 gigawatts, the EPS uplift could reach about 10%. Strategic motivations: Insufficient investments from hyperscale cloud providers, NVIDIA Corporation seeks new growth opportunities. The Morgan Stanley report reveals the strategic logic behind NVIDIA Corporation's move. NVIDIA's management believes that traditional hyperscale cloud providers are constrained by pressures on free cash flow and a growing scarcity of land, power, and IT resources within the U.S., making it difficult for their capital expenditure expansion to keep pace with the increasing demand for tokens, and they are even being forced to reallocate computing power originally meant for model training to inference scenarios. At the same time, GEO Group Inc's political factors are accelerating this evolution. As U.S. cloud providers face policy pressures when building data centers in other countries, some nations and regions have begun to view local land and power resources as strategic assets for AI, instead supporting local sovereign clouds or regional neocloud providers. Morgan Stanley cites Yotta, YTL, and SK Telecom as typical examples. Against this backdrop, NVIDIA Corporation has chosen to intervene proactively, cultivating a new group of computing power buyers through capital endorsements, thereby bypassing hyperscale cloud providers' capital expenditure constraints while reinforcing its dominance in the global computing power landscape. The report indicates that the current capital expenditure scale of publicly listed neoclouds already accounts for about 6% of global cloud capital expenditure, with significant growth potential, not including sovereign and private cloud players. Existing Layout: From CoreWeave to a $500 Billion Industry Fund. NVIDIA Corporation has long been positioned in the neocloud sector. According to the Morgan Stanley report, since investing in CoreWeave in 2023, NVIDIA has extended this strategy to multiple cloud providers: currently holding over 10% equity in CoreWeave and pledging to purchase all unsold computing power until April 2032; holding approximately 10% equity in Nebius; signing a $3.4 billion, five-year cloud service contract with IREN, which includes a $2.1 billion warrant; and also engaging in private equity investments in Nscale, Crusoe, Lambda, Firmus, and others. On the software front, NVIDIA Corporation has launched the DGX Cloud Lepton platform, which serves as a market-oriented matching platform connecting developers with the GPU computing power of its cloud partners, dynamically routing computing demands to cloud providers with available capacity. In terms of hardware infrastructure, NVIDIA Corporation has recently taken substantial steps. Reports indicate that NVIDIA Corporation, in partnership, has announced the construction of 16 new fiber optic lines in the U.S., involving 8,000 miles of long-haul fiber, with neocloud support explicitly listed as one of the project drivers. Additionally, it has been reported that NVIDIA Corporation is signing a 15-year, $20 billion lease with HUT 8 for a 1-gigawatt Beacon Point facility. In terms of capital operations, NVIDIA Corporation recently announced the mobilization of a $50 billion investment fund, with an official statement clearly stating that the fund will be used to "support long-term usage-linked earnings." Morgan Stanley points out that the funds mechanism requires NVIDIA Corporation to control its maximum direct exposure to any single investment to within 25%, thus somewhat mitigating cyclical risks. Bull and Bear Combat: High Growth Potential Coupled with Cyclical Risks. Morgan Stanley acknowledges that the new model has sparked significant divergence in the market. Bulls argue: First, the current supply of computing power continues to lag behind the growth in token demand, with new capacity supported by real demand; second, securing annuity-like income through minority equity stakes in multiple cloud service providers will enhance the predictability of NVIDIA Corporation's long-term profitability; third, this move will help NVIDIA further establish its chips as the de facto industry standard as the new product cycle of Vera Rubin approachesadopting NVIDIA chips, which hold approximately 85% market share and have the most complete ecosystem, remains the optimal choice for smaller cloud providers. Bears, on the other hand, focus on the following risks: this model may create "artificial demand"for instance, without NVIDIA Corporation's credit endorsement, some computing power investments would not materialize; furthermore, the capital strength of participants varies, and the overall model may face significant pressure in a scenario of computing power surplus. Morgan Stanley states that high transparency in disclosures is key to maintaining market confidence in this model. In a bearish scenario, Morgan Stanley also points out a potential hedge path: if there is an oversupply of computing power, NVIDIA Corporation could utilize the GPU usage rights obtained to advance its own model development. Currently, NVIDIA's flagship open-source model, Nemotron 3 Ultra, boasts 550 billion parameters and is leading the market outside of China; the next-generation Nemotron 4 is expected to surpass the trillion-parameter threshold, further strengthening its competitiveness in model development. Impact on Existing Neocloud: Intensified Competition, Execution Capability is Key. Morgan Stanley points out that NVIDIA Corporation's move is a double-edged sword for existing neocloud players. In the long term, the increase in potential competitors poses pressure; however, in the short term, reliable operators among cloud service providers have the opportunity to take on more new computing power investments. The report emphasizes that NVIDIA Corporation's credit endorsement merely addresses financing issues; cloud service providers still need to complete construction on time and convert it into revenue within the complex geopolitical and supply chain environment posed by GEO Group Incplacing high demands on operational execution capabilities. Overall, Morgan Stanley maintains NVIDIA Corporation as a preferred pick in the semiconductor industry, with a target price of $288, corresponding to approximately 22 times the CY27 P/E ratio valuation. Analysts expect NVIDIA Corporation's revenue to grow by 82% year-over-year in FY2026, and to further increase by 52.4% in FY2027; FY29 EPS is forecasted at $17.63. This article is reproduced from "Wall Street Insights," author: Zhang Yaqi; GMTEight editor: Yan Wencai.