Bullish has provided USD.AI with a $100 million stablecoin debt facility to fund loans secured by graphics processing units used in artificial intelligence infrastructure.
Key Takeaways
Bullish supplied $100 million in stablecoin liquidity for GPU-backed loans
USD.AI offers non-recourse financing secured solely by computing hardware
Bullish plans to list sUSDai yield-bearing token with a market-making program
Existing USD.AI loans include 3,072 GPUs and over $132 million in financing
Bullish Facility Expands USD.AI's GPU Loan Capacity
Bullish said the facility will give USD.AI capital to finance operators purchasing high-performance computing equipment, extending the exchange operator's exposure to tokenized assets and AI infrastructure. Developed by Permian Labs, USD.AI connects stablecoin liquidity with companies seeking funds to buy GPUs. Instead of evaluating claims against a borrower's entire business, the platform issues non-recourse loans secured by the computing hardware purchased with the financing.
Our commitment to USD.AI reflects a conviction we've believed since our first investment in the protocol: that credible, well-structured real-world assets belong onchain.
— Thomas Cowan, Head of Tokenization at Bullish
Under the arrangement, USD.AI will use the $100 million facility to originate loans for middle-market AI infrastructure operators. Ownership of the financed hardware provides the collateral, while the loan does not create a claim against the operator's other corporate assets. This connects to the broader RWA and AI infrastructure trend covered in Mantle's $880M in stablecoins and tokenized assets and the Stellar's $3B RWA market.
USD.AI's Existing GPU Loans
USD.AI has already completed large transactions involving recent Nvidia hardware. In June, the protocol announced a $98.1 million loan backed by 2,304 Nvidia B300 GPUs. Investors fully funded another $34 million facility secured by 768 Nvidia B200 units. Combined, the two disclosed loans involved 3,072 GPUs and more than $132 million in financing. The latest Bullish facility gives the protocol another source of stablecoin liquidity as it adds loans for operators building data centers and AI computing clusters.
Permian Labs CEO David Choi said demand for computing equipment has created a distinct lending category. Compute is becoming a credit market in its own right, Choi said, adding that Bullish's facility would allow USD.AI to finance more infrastructure and develop trading markets for compute-backed debt. This connects to the NVIDIA revenue hitting $96.2B and broader AI-crypto convergence.
sUSDai Listing Creates Secondary Trading Market
Alongside the lending facility, Bullish plans to introduce sUSDai across several trading pairs on its institutional exchange. A dedicated market-making program will provide orders for the token once trading begins. USD.AI uses sUSDai as its yield-bearing token, giving holders exposure to returns generated from the protocol's credit operations. Listing the asset would allow holders to trade the position instead of relying only on the repayment period of underlying loans.
Bullish expects the program to improve secondary liquidity and price discovery for GPU-backed debt. The companies did not disclose planned trading pairs, launch date, or market-making budget. Tokenized exposure to computing hardware has appeared elsewhere, including Injective's Nvidia GPU derivatives market and Aethir's tokenized GPU marketplace.
Bullish Deepens Existing USD.AI Relationship
The $100 million facility follows Bullish Capital's $4 million investment in USD.AI in September 2025. Bullish operates under the EU's Markets in Crypto-Assets framework as an authorized crypto asset service provider and obtained a New York BitLicense in September 2025. Bullish completed its NYSE debut in August 2025, pricing shares at $37 and raising about $1.03 billion. Despite recent recovery, BLSH remains more than 60% below its $90 opening price, trading around $33 after gaining approximately 45% over the past month.













































