- A new stablecoin protocol converts crypto liquidity into loans for GPUs used in Artificial Intelligence (AI).
- The protocol’s synthetic dollar, USD.AI, is backed by credit tied to NVIDIA GPUs in rented data centers.
- Loans use tokenized GPU receipts as collateral, allowing on-chain capital to fund physical AI hardware.
- Lenders earn yields from GPU operators’ repayments, currently ranging from 13% to 17%.
- The system’s design aims to decentralize infrastructure financing, potentially extending beyond AI.
A digital finance protocol called USD.AI transforms idle cryptocurrency funds into loans for the GPUs that power artificial intelligence. The system, which currently has approximately $345 million in circulation, supports its synthetic dollar through short-term credit linked to NVIDIA GPUs housed in data centers rented to AI developers.
The GPUs generate income by selling computation time needed for AI model training and inference. This income repays the loans, allowing lenders to earn returns from real economic activity instead of token rewards. Borrowers receive specialized financing that traditional retail lenders usually avoid due to higher risk.
The protocol uses three main mechanisms. First, CALIBER creates a legal and technical connection between physical GPUs and their on-chain representation. Each GPU is kept in an insured data center and represented by a non-fungible token (NFT) that legally claims ownership of the hardware. Loans are then issued using these tokenized assets as collateral.
Second, the FiLo Curator functions as an underwriter. Curators originate and manage loans, putting up their own funds to cover initial losses. This method decentralizes credit approval while aligning incentives since curators profit only if borrowers repay.
Third, QEV (queue extractable value) manages liquidity. Instead of allowing instant withdrawal of funds, redemption requests enter a queue. Users who wait receive repayments gradually from borrower repayments, while those who want faster exits can pay a fee to move ahead in line. This arrangement protects the loan book’s solvency and rewards patient lenders.
Currently, staking USD.AI’s stablecoin yields between 13% and 17%, with repayments coming from GPUs rather than inflationary token issuance. Supporters view USD.AI as a prototype for “InfraFi,” decentralized infrastructure finance, which may eventually cover areas like renewable energy or decentralized computing networks.
The future success of USD.AI depends on sustained demand for GPU leasing, reflecting ongoing needs in AI development. If that demand holds, this protocol could provide a significant link between on-chain capital and the physical machines driving artificial intelligence.
For a detailed view of USD.AI’s circulation, see the Dune Dashboard. Additional information on the protocol’s launch is available at this update.
✅ Follow BITNEWSBOT on Telegram, Facebook, LinkedIn, X.com, and Google News for instant updates.
Previous Articles:
- GlassWorm Worm Targets VS Code Extensions in Major Supply Chain Attack
- BRICS Experiences Unprecedented Gold Boom with India and China Discovering Record-Breaking Mines
- Trump Pardon Bet Odds Jump for Bankman-Fried, Ver After Zhao Release
- US Energy Chief Urges FERC to Fast-Track Big Power Users Access
- WazirX Set to Restart Cryptocurrency Trading Officially
