Nvidia has partnered with six prominent Wall Street firms to promote "AI compute" as a viable infrastructure asset.
By Omkar Godbole, AI Boost|Edited by Shaurya Malwa55 min ago4 min readMake preferred on ShareShare this articleCopy linkX (Twitter)LinkedInFacebookEmailMake preferred on Nvidia aims to make AI chips an investable asset class. (Possessed Photography/Unsplash)SummaryShow- Nvidia has entered agreements with six major Wall Street firms to develop funding platforms that could channel over $500 billion into AI computing infrastructure.
- The company seeks to classify AI compute, primarily driven by its GPUs in "AI factories," as long-term revenue-generating infrastructure instead of short-lived tech costs.
- This initiative may further widen the gap with smaller decentralized computing networks that face technical and operational limitations, keeping their capabilities significantly lower than advanced data centers.
Nvidia, a key player in the AI sector, is advocating for financial institutions to recognize its AI computing capabilities as similar to commercial real estate, toll roads, or power plants—viewing them as investable infrastructure assets.
On Monday, Nvidia announced it had signed memorandums of understanding with six influential Wall Street firms: Apollo Global Management, Blackstone, BlackRock, Brookfield Asset Management, Goldman Sachs, and KKR. These partnerships aim to create financing platforms that could potentially access over $500 billion in external capital.
The chipmaker's objective is to redefine AI compute as a bankable infrastructure asset, encouraging clients to establish AI data centers and secure ongoing demand for Nvidia's hardware.
Jensen Huang, Nvidia's founder and CEO, remarked, "This is really the first time that technology chips have become an investable asset class. These are revenue-generating assets now. They’re productive, they’re long-lived, they’re fungible, they’re flexible." He further mentioned, "Fundamentally, what’s different about this industry and this way of doing computing is that the computer is now part of the infrastructure, like electricity, like the internet, and so you have to think about it like it’s infrastructure."
Understanding AI Compute
AI compute refers to the computational power utilized for training and executing artificial intelligence models, predominantly reliant on specialized chips, mainly Nvidia's high-performance GPUs, which form the backbone of what the company describes as "AI factories."
These factories leverage electricity and data to produce sophisticated intelligent systems, such as those that enable chatbots, generate multimedia content, assist in drug design, control robotics, and facilitate numerous other applications. The complexity of the AI dictates the demand for this specialized computing power.
Nvidia's Vision for Change
Currently, many companies perceive the acquisition or leasing of computing power as a typical technological expense that depreciates quickly as newer models emerge. (Think of how quickly smartphones become outdated.)
However, Nvidia argues that this perspective is outdated, asserting that its systems are widely adopted and can serve multiple customers, thus generating a revenue stream over several years. Consequently, AI factories should be regarded as long-term investable assets.
To illustrate the shift Nvidia is advocating: Suppose a company currently requires advanced AI chips. It would typically invest millions of dollars from its own funds or secure a conventional business loan to procure a large array of Nvidia GPUs. This expense is recorded on the balance sheet as equipment, which accountants depreciate over time as newer models become available. The expectation is that before the equipment loses substantial value, the AI-driven product will generate enough revenue to surpass the initial investment.
Now, under Nvidia's proposed model, the company would still need the chips but wouldn't have to pay the full price upfront. Instead, they would utilize a financing platform supported by major institutional investors. In return, these investors would earn revenue from the rental income generated by the chips over many years.
This approach allows companies to access the processing power of AI factories through rental payments. Institutional investors find this appealing because the same Nvidia chips can cater to various clients and tasks, creating a consistent income stream.
This transformation is central to the announcements made on Monday.
According to the MoUs, Wall Street banks will evaluate each project for customer demand, anticipated usage, and cash flow before allocating funds. In some instances, Nvidia may assume 25% of the risk if the chips depreciate in value, but the key point is that lenders will conduct their own assessments and make independent decisions regarding each project.
Nvidia's initiative could potentially unlock a broader pool of long-term capital at a time when investors are increasingly questioning the potential returns on the substantial capital expenditures by major tech companies in AI.