
SENTORA WEBINARS
About the Webinar
AI runs on GPUs, data centres, power and cooling, and every one of them must be financed before it earns revenue.
The scale of that financing is growing fast. In 2026, compute deals ranged from multi-billion-dollar GPU loan facilities to GPU finance leases of up to $40 billion, drawing in sponsors, private credit funds, banks and hardware vendors. These deals layer sponsor equity, mezzanine capital and senior debt, and each layer prices risk differently.
Onchain markets, with their programmable settlement, transparent collateral data and global investor access, could become a meaningful channel for this capital. Getting there takes more than tokenization. Putting a GPU or a lease onchain makes it easier to hold and transfer, while financing it requires underwriting, structuring and active risk management over the life of the loan.
This webinar examines how today’s largest compute financings are structured and what an onchain version of that capital stack would need to look like.
We will discuss the four risks that decide GPU-backed credit: duration mismatch between loan terms and hardware life, the strength of the contracts behind the cash flows, obsolescence as new chip generations arrive, and concentration among a small group of borrowers and suppliers.
Sentora’s perspective builds on its work in onchain credit, risk curation, and AI research. The objective is a clearer view of where onchain finance offers real potential for AI infrastructure today, the hurdles it must overcome, and what builders and allocators should consider as these two markets converge.

