# Issuance and Redemption

USDAI’s lifecycle—how it’s created and retired—is designed to be user-friendly yet secure, integrating seamlessly with GPU.NET’s ecosystem.

#### Minting

* **Process**: Users can **mint USDAI** by locking compute credits or depositing supported assets into GPU.NET’s smart contracts. Supported assets include:
  * **Compute Credits**: Providers contribute GPU capacity directly, receiving USDAI proportional to the market value of their contribution.
  * **Cryptocurrencies**: Assets like ETH, SOL, or other stablecoins (e.g., USDC) can be deposited, with the smart contract converting their value into USDAI based on the $1 compute peg.
* **Example**: A provider locks 50 GPU-hours (valued at $50 USD) into Dapp.gpu.net. The system mints 50 USDAI, which the provider can use or trade. Alternatively, a user deposits 0.02 ETH (worth $50 at current rates), receiving 50 USDAI in return.
* **Fees**: Minting incurs a small network fee (e.g., 0.5%) to cover operational costs and incentivize validators.

#### Burning

* **Process**: USDAI can be **burned** (removed from circulation) to redeem compute credits or withdraw other assets:
  * **Compute Redemption**: Users burn USDAI to unlock GPU capacity on Dapp.gpu.net for immediate or scheduled use.
  * **Asset Withdrawal**: USDAI can be swapped back into supported cryptocurrencies, subject to liquidity and network availability.
* **Example**: A developer burns 10 USDAI to access 10 GPU-hours for ML training. Alternatively, they burn 10 USDAI to withdraw $10 worth of SOL, minus a redemption fee.
* **Availability**: Redemptions depend on compute liquidity and asset reserves, managed by Grid.gpu.net’s validators.