Nvidia-backed Zankore launches US$3.1 billion GPU loan
The Indonesian AI infrastructure firm is seeking major debt financing to purchase the hardware needed for its new AI factory.
The short version
- Zankore is seeking a US$3.1 billion (RM12.65 billion) loan to purchase GPUs for a new AI factory.
- The five-year loan carries an interest margin of 330 basis points over the Secured Overnight Financing Rate.
- Nvidia provides credit support and revenue sharing for the deal, covering unsold compute capacity from an initial 100MW.
- The loan is being marketed by five original underwriters, including United Overseas Bank and Citigroup.
- US fintech companies are the primary customers purchasing the compute capacity from Zankore.
Why it matters
This deal signals a shift in how lenders view AI infrastructure, moving from just funding buildings to financing the actual chips inside them. It also shows how hardware giants like Nvidia are increasingly involved in the financial structures of their customers.
The financing structure
Zankore, an Indonesian platform for artificial intelligence infrastructure, is moving to secure a massive amount of debt to fuel its expansion. As reported by The Edge Malaysia, the firm is syndicating a US$3.1 billion loan to the broader financial market. This capital is specifically earmarked for the purchase of graphics processing units, which are the essential chips used to train and run large AI models.
The loan is structured as a five-year agreement with an interest margin of 330 basis points over the Secured Overnight Financing Rate. This rate is a standard benchmark used to determine the cost of borrowing money overnight. By using a floating rate, the loan costs will adjust based on broader market conditions over the life of the debt.
To market this deal, Zankore has engaged five original underwriters, which include major financial institutions like United Overseas Bank and Citigroup. Other banks have also been invited to participate in the senior phase of the loan as bookrunners or underwriters. This high level of institutional involvement suggests a significant appetite for AI-related debt in the current market.
Nvidia's role in the deal
A critical component of this financing is the involvement of Nvidia, the dominant manufacturer of AI chips. Zankore is backed by Nvidia, and the firm has secured a six-year contract with the hardware giant to support this specific expansion. This contract includes provisions for revenue sharing and credit support, which are designed to make the loan more attractive to cautious lenders.
The credit support from Nvidia is particularly important because it helps reduce the risk of default for the banks providing the money. The arrangement specifically applies to any compute capacity that remains unsold from an initial 100MW output. By providing a safety net for unsold capacity, Nvidia ensures that the infrastructure remains viable even if demand fluctuates in the short term.
This partnership highlights a growing trend where hardware providers are becoming deeply integrated into the business models of their clients. Rather than just selling a product, Nvidia is helping to structure the financial backbone of the infrastructure that uses its products. This move helps stabilize the investment for the banks while ensuring Zankore has the hardware it needs to operate.
The demand for compute
The primary customers for Zankore’s new AI factory are expected to be US fintech companies. These companies are the offtakers, meaning they are the ones who will actually buy and use the GPU compute capacity provided by the Indonesian platform. This indicates a strong cross-border demand for high-performance computing power originating from the United States.
The deal comes at a time when regional lenders in Asia are looking for new ways to participate in the AI boom. While many have historically focused on the construction of data centers, there is a growing shift toward financing the actual hardware inside those buildings. This transition allows lenders to capture a larger piece of the value chain associated with AI development.
The scale of this investment reflects the broader expectations for the industry's growth. Even as some investors remain cautious about high valuations in the AI space, the underlying demand for physical infrastructure remains robust. Zankore’s move to secure billions for GPUs suggests that the
Broader market trends
Zankore’s loan is part of a wider wave of debt financing across the Asian region. Other major infrastructure projects are also seeking significant funding to keep up with the demand for AI capabilities. For example, Blackstone-owned AirTrunk recently secured a A$4.25 billion green loan for a data center campus in Sydney.
Similarly, Stack Infrastructure Inc, which is owned by Blue Owl Capital, is currently seeking a A$8.6 billion facility to build out a campus in Melbourne. These multi-billion dollar deals show that the capital required to build the backbone of the AI economy is reaching unprecedented levels. Investors are increasingly looking for diverse ways to fund these massive physical requirements.
Moody’s Ratings has provided a broader outlook on this trend, expecting at least US$3 trillion in data center-related investment over the next five years. This massive figure underscores the scale of the buildout that is currently underway. As more firms like Zankore enter the market, the competition for high-quality debt and hardware will likely intensify.
What comes next
The successful closing of Zankore's loan will provide the necessary capital to begin the heavy lifting of hardware procurement. Once the GPUs are secured and the factory is operational, the focus will shift to managing the 100MW of compute capacity. The firm will then work to fulfill the contracts with the US fintech companies that are eager to access this power.
For the financial markets, this deal serves as a blueprint for how other infrastructure companies might seek funding in the future. By bundling hardware purchases with manufacturer support and revenue-sharing agreements, companies can mitigate the risks that often come with high-tech investments. This could lead to more standardized ways of financing AI infrastructure across the globe.
Observers will be watching to see how these large-scale GPU loans perform as the AI market matures. If the demand from offtakers remains steady, it will validate the current lending models. If demand fluctuates, the credit support from Nvidia will be put to the test, determining how well these complex financial structures hold up under pressure.
What to watch
- Whether other Asian infrastructure firms follow Zankore's model of using manufacturer credit support to secure debt.
- The actual utilization rate of the 100MW compute capacity by US fintech companies.
Try it
Look up the current Secured Overnight Financing Rate (SOFR) to see the benchmark used for this loan's interest calculation.
Sources and links
Liked it? Say why
Add a line for other readers.
No account needed. We never show your email. Comments with links, ads or abuse are held and not shown. Your comment shows after a quick check.
Listen to the AI Wire headlines
Listen to the AI WireListen to our daily AI shows every morning: on your commute, in the shower, wherever you are
Listen to the daily showsFirst reported by The Edge Malaysia · Read the original. Written by the SI Agent News Desk from the reporting above.Written with AI from the source above. Every number is checked against the source. Spot a mistake? Corrections.
