Solution
GAIB solves the problems by building a financial infrastructure for AI, starting with a new financial primitive for AI compute. GAIB tokenizes GPUs and their cash flows to create a new type of assets, and builds a DeFi ecosystem on top. This approach not only accelerates the expansion of cloud and data centers by facilitating funding, but also offers investors direct exposure to compute assets and their products, giving everyone equitable access to dividends of the AI era.
Off-Chain Yield Sources: GPU Financing
GAIB structures financing deals with established cloud operators and data centers worldwide. These deals come in three main forms: a debt model that takes interest on borrowed capital; an equity model that bases repayment on a share of GPU-generated revenue; and a hybrid model combining a blend of both. All agreements are backed by enterprise-grade GPUs (and their service contracts) — including NVIDIA H100s, H200s, GB200s and more — which are widely sought after and often reserved by clients for one to three years.
By using a bankruptcy-remote structure and prioritizing well-secured clusters, GAIB can offer financing maturities as short as three months (for bridge financing) or up to three years. Thanks to the high demand for these GPUs, the timeline for repayment can be much shorter than typical debt or equity deals. The projected gross yield for each deal varies depending on the financing model: around 10–20% for debt arrangements and up to 60–80% or more in equity-based structures.
Stringent due diligence underpins every deal. GAIB and other partnered underwriters on the platform follow a robust Credit Analysis framework, often collaborating with third-party auditors when needed. Each loan is over-collateralized by the physical GPUs and their associated service contracts. In case of default, the underwriter is empowered to liquidate the GPUs to protect investors or to continue hosting and managing them through strategic data center partners, ensuring ongoing revenue generation.
Given the surge of decentralized compute protocols aggregating underutilized compute resources and bridging compute supply and demand via blockchain, GAIB partners up with them to tokenize the on-chain revenue streams of compute providers on different de-compute platforms, starting with Aethir. Details can be found here.
Bridging Off-Chain and On-Chain: The Power of Tokenization
GAIB unites real-world AI compute with decentralized finance by creating a tokenized portfolio of GPU financing deals, drawn from both off-chain data centers and on-chain partners. A small reserve of treasury bills is also included for liquidity management. This diversified portfolio backs GAIB’s AI Synthetic Dollar, known as AID, seamlessly linking the yield potential of physical compute assets with the efficiency of on-chain markets.
On-Chain Innovation: AI Synthetic Dollar “AID”
Because AI compute captures the lion’s share of value in the emerging AI landscape, GAIB’s tokenization strategy centers on turning compute assets and their yields into a cryptocurrency. AID stands as the first token designed to fuse AI’s economic potential with liquidity, making it easy to buy and sell through various Automated Market Maker (AMM) pools. Whitelisted partners who have completed KYC can also mint and redeem AID directly with GAIB.
GAIB’s commitment to building a vibrant DeFi ecosystem sets it apart from incumbent Real-World Asset (RWA) protocols. Beyond the high yields, relatively shorter terms, and robust liquidity of GPU-backed deals as the underlying assets, GAIB offers a wide spectrum of financial products and strategies via its own platform and integrations with major DeFi protocols. Holders of AID can stake to earn a native yield, lend for additional returns, trade specialized principal or yield tokens (PT or YT) derived from staked AID (sAID), or create customized yield strategies based on AID. By building a range of use cases for AID and sAID, GAIB aims to serve users with all risk-reward profiles.
A Currency for the AI Era
Looking ahead, GAIB plans to build more financial primitives across the AI stack—including agents, models, and data—which will open new use cases for AID. The overarching goal is for AID to evolve into a base currency that facilitates, stabilizes, and democratizes access to the vast potential of AI. By weaving real-world compute assets into the on-chain economy, GAIB is not only accelerating the growth of AI infrastructure but also creating a new paradigm where anyone can share in the dividends of the AI revolution.
Last updated