Nvidia, CoreWeave, and Nebius: Inside the Circular Financing of the GPU Boom
370 points
• 5 days ago
• Article
Link
Neocloud 公司，例如 CoreWeave 和 Nebius，已成为 AI 基础设施领域的核心玩家，拿下大型 hyperscaler 的合作并签订了数千兆瓦的电力合同。它们对 Microsoft 和 Meta 等科技巨头的吸引力在于能够快速提供最新的 Nvidia GPU 硬件并提高计算利用率。通过将基础设施需求外包给 neocloud，hyperscaler 可以把巨额资本支出转为运营支出，在扩展 AI 能力的同时保护自身资产负债表。
尽管收入增长迅猛，这些 neocloud 的风险远高于成熟科技公司。两家公司都面临着巨额资本支出与有限现金流之间的严重错配。为支撑激进扩张，它们背负了大量债务——尤其是 CoreWeave，在把已签约的电力转化为可用产能的过程中债务负担尤重。尽管它们主张专有软件和基础设施优化带来竞争优势，但对外部融资的高度依赖仍是重大隐忧。
推动这种增长的一个关键因素是 Nvidia，它既是供应商也是金融后盾。 Nvidia 向这些公司投入了数十亿美元的股权资金，同时承诺购买任何剩余的、未售出的 GPU 容量。这形成了一种循环融资机制：Nvidia 在一定程度上资助了购买其芯片的实体。虽然这一关系为 Nvidia 锁定了长期销量并助力 neocloud 扩张，但如果这些公司无法在没有 Nvidia 支持的情况下实现独立盈利，该模式的长期可持续性就值得怀疑。
利率上升等外部经济因素进一步加剧了它们的财务压力。利息支出已吞噬收入的很大一部分，使这些公司对宏观波动高度敏感。随着它们继续依赖债务工具填补资金缺口，资本成本可能变得愈发沉重，进一步压缩利润空间。
总之，neocloud 在当前的 AI 热潮中举足轻重，但它们的未来取决于能否摆脱这些以债务驱动的复杂融资模式，转而建立自我维持、现金流为正的商业模式。
Neocloud companies like CoreWeave and Nebius have become central players in the AI infrastructure landscape, securing massive hyperscaler partnerships and multi-gigawatt power contracts. Their appeal to tech giants like Microsoft and Meta lies in their ability to provide rapid access to the latest Nvidia GPU hardware and superior compute utilization. By offloading these infrastructure needs to neoclouds, hyperscalers can effectively move enormous capital expenditures into operating expenses, shielding their own balance sheets while still scaling their AI capabilities.
Despite their rapid revenue growth, these neoclouds operate with significantly higher risk profiles compared to established tech firms. Both companies face a daunting mismatch between their massive capital spending requirements and their actual cash flow. They have incurred substantial debt to fund their aggressive expansion, with CoreWeave in particular carrying a heavy debt load as it attempts to convert contracted power into active capacity. While they argue their specialized software and infrastructure optimizations offer a competitive edge, their reliance on outside funding to sustain these capital-intensive operations remains a significant concern.
A critical aspect of this growth is the role of Nvidia, which acts as both a supplier and a financial backstop. Nvidia has made multi-billion-dollar equity investments in these firms while simultaneously providing financial guarantees to purchase any residual, unsold GPU capacity. This arrangement creates a circular financing structure where Nvidia effectively helps fund the entities that are purchasing their own chips. While this relationship secures Nvidia's long-term sales and helps the neoclouds scale, it raises questions about the long-term sustainability of the model if these companies cannot achieve profitability independently of Nvidia's support.
External economic factors, such as rising interest rates, further complicate the financial outlook for these companies. With interest expenses already consuming a significant portion of revenue, these firms remain highly sensitive to fluctuations in the macro environment. As they continue to draw on debt facilities to bridge their funding gaps, the cost of capital could become an increasingly heavy burden, placing further pressure on their bottom lines. Ultimately, while neoclouds are currently integral to the AI boom, their future depends on their ability to transition from these complex, debt-fueled financing models to a self-sustaining, cash-flow-positive business.
180 comments • Comments Link
• Nvidia 在 "Neoclouds" 的投资是对 Hyperscalers 的一种战略对冲，既确保了 Nvidia 全栈技术（包括网络与存储）的部署，也为合作伙伴提供早期访问并为 Nvidia 带来宝贵的使用数据。
• Nvidia 放弃通过 DGX Cloud 与 Hyperscalers 直接竞争，转而采取这种投资合作方式，从而在实现类似基础设施目标的同时，不会疏远那些优先考虑自研芯片的核心客户。
• 批评者认为，这种融资模式存在隐患：它通过长期合约杠杆化债务，形成循环依赖——收入增长可能依赖于由供应商资助的客户，一旦需求或流动性枯竭便可能出现严重问题。
• 交易结构透明度不足，例如使用 SPVs 促成 GPU 采购，这使得难以判断所报告的收入是真实市场需求还是人为放大的循环资本。
• 目前难以评估这些建设项目的经济可行性，行业健康最终取决于 ROI per token 、企业预算的可持续性，以及在硬件快速淘汰周期中保持定价权的能力。
• 硬件效率提升速度极快，这对基础设施提供商构成重大风险：面对更高效的新架构，旧的昂贵 GPU 集群可能失去经济竞争力，从而导致大范围资产贬值。
• 市场参与者对于这是否只是标准泡沫还是系统性金融风险存在分歧；有人将当前状况比作 1929 年的崩盘，认为过度投资与 "too big to fail" 的实体可能隐藏波及更广泛经济的脆弱性。
• 一些观察者指出，受电力与许可等制约，数据中心部署步伐放缓，这在某种程度上起到了自然制动作用，可能防止 AI 投资泡沫破裂后出现的大规模产能过剩。
• 关于 "circular financing" 是否构成有效批评存在争议：有人认为所有企业融资本质上都有循环性，另一些人则坚持认为将投资与收入混为一谈是会计操纵的危险信号。
• 由于交易条款很少全面披露，围绕这些 AI 基础设施投资的不透明性，仍然是投资者怀疑与对未来潜在减记的主要担忧来源。
这场讨论反映出人们对 Nvidia 当前投资策略看法的深刻分歧：一方将其视为硬件密集型市场中的合理防御举措，另一方则怀疑资本流动依赖于不透明且可能人为夸大需求的循环机制。多头强调控制生态系统与获取数据的战略必要性，空头则强调资产淘汰风险以及涉及客户融资模式的会计欺诈历史先例。归根结底，由于缺乏细粒度的财务数据，AI 基础设施繁荣的长期可持续性仍未可知，目前唯一的共识是：行业发展之快使得短期预测高度投机性。 • Nvidia's investments in "Neoclouds" act as a strategic hedge against hyperscalers, ensuring deployment of Nvidia's full stack—including networking and storage—while securing early access for partners and providing Nvidia with valuable usage data.
• Direct competition with hyperscalers via DGX Cloud was abandoned in favor of these investments, which accomplish similar infrastructure goals without alienating major customers who would otherwise prioritize their own proprietary chip designs.
• Critics argue the financing model is potentially problematic because it uses long-term contracts to leverage debt, creating a circular dependency where revenue growth may rely on customers funded by the supplier, raising questions about what happens if demand or liquidity dries up.
• The lack of transparency in deal structures, such as the use of Special Purpose Vehicles (SPVs) to facilitate GPU purchases, obscures whether reported revenue represents genuine market demand or artificially inflated, recycled capital.
• Assessing the economic viability of these builds is currently difficult, but industry health will eventually hinge on ROI per token, enterprise budget sustainability, and the ability to maintain pricing power as hardware faces rapid obsolescence cycles.
• Hardware efficiency improvements are occurring so quickly that a significant risk exists for infrastructure providers: older, expensive GPU fleets may become economically uncompetitive against newer, more efficient architectures, potentially leading to widespread asset devaluation.
• Market participants are divided on whether this constitutes a standard bubble or a systemic financial risk, with some comparing current conditions to the 1929 market crash where over-invested, "too big to fail" entities hide underlying fragility that could ripple through the broader economy.
• Some observers suggest the slow pace of datacenter rollouts, constrained by power and permitting issues, might act as a natural brake that prevents the massive surplus capacity that would otherwise follow an AI investment bubble burst.
• There is a debate over whether the term "circular financing" is a valid critique, as some argue all corporate financing is inherently circular, while others maintain that specifically conflating investment with revenue is a historical red flag for accounting manipulation.
• Because full disclosure of deal terms is rare, the opacity surrounding these AI-related infrastructure investments remains the primary source of investor skepticism and concern regarding the potential for future write-downs.
The discussion reflects a deep split between those who view Nvidia's current investment strategy as a rational, defensive play in a hardware-intensive market and those who suspect the capital flow relies on opaque, circular mechanisms that artificially inflate demand. While bulls focus on the strategic necessity of controlling the ecosystem and gaining data, bears emphasize the risk of asset obsolescence and the historical precedents of accounting fraud involving customer-funding models. Ultimately, the lack of granular data on these financial arrangements leaves the long-term sustainability of the AI infrastructure boom as an open question, with consensus appearing only on the fact that the industry's rapid pace makes near-term predictions highly speculative.