If you notice @USDai_Official is solving a real problem in AI infra, which is funding bottlenecks. TradFi is too slow for DePIN companies. Loan approvals take months. GPU shortages exist not because of supply, but because of capital access issues. Let us understand what it is tackling🧵
1/ Brookfield projects that total AI data center capacity will expand by more than 10-fold between 2024 and 2034 (rising from 7 GW to 82 GW) Making AI infra spending hit $7T over the next decade. $4T for GPUs alone. But getting funding through banks is slow & expensive. GPU as a Service market is projected to grow from $8-30B now to $250-500B by 2034.
In September, @USDai_Official's TVL surged 350%, driven by deposit cap increases that expanded participation in their ICO allocation/airdrop program called “The Allo Game.” Over the course of the month, deposit caps were raised multiple times, from an initial $110 million cap up to $500 million on September 23. In all cases, the new cap was filled within a day. Despite this latent demand being tied to incentives, I believe it's likely that growth remains strong long after the program ends. USDAI is at the right place at the right time, sitting at the intersection of two high-growth markets and uniquely benefiting from tailwinds in each. Compute Tailwinds: The “GPU-as-a-service” market is forecasted to grow by as much as 4-5x in the next 10 years, driven by a rising demand for inference. While training accounts for ~80% of compute demand today, many firms project that inference will represent 50-75% by 2030. Inference can be served by smaller players, and as the compute landscape shifts, USDAI’s model offers these players an edge in capital efficiency. Stablecoin Tailwinds: USDAI also stands to benefit from the anticipated Fed rate-cut cycle. Rate cuts will compress the Treasury-based yields that stablecoin issuers rely on. As those yields decline, capital may shift toward stablecoins offering alternative, higher-yield mechanisms. For more info on USDAI's positioning and adoption, as well as other sections from gigabrain analysts, check out the Messari monthly report below 👇
2/ @USDai_Official created USDai which is a synthetic stablecoin backed by two things.. > US Treasury Bills (4% yield) > Tokenized GPU hardware leased to AI companies (15-20% yield) Currently the TVL is $578M. Current APY is ~18.2% for stakers. Nearly 72% of it is on Arbitrum while the remaining is on Plasma. h/t to @EntropyAdvisors @Dune for the data✍️
4/ For depositors you mint USDai with USDC or USDT. Stake it as sUSDai for yields. The peg stays at $1 through overcollateralization (over 150%). Capital flows to T-Bills & hardware loans. You earn from both.
5/ The CALIBER framework handles transparency. All assets tokenized onchain. If someone defaults, their hardware NFT gets auctioned automatically. Proceeds go back to the system. This removes offchain trust issues that plague traditional lending.
6/ Most liquidity is on Pendle with small amounts on Curve, Uniswap, Euler & Fluid. Recent integrations with Aave passed approval where USDai can be used as collateral there now. h/t to @Jonasoeth @Dune for the data✍️
8/ @arbitrum wins in this as it makes the entire model economically viable. Transaction costs on Ethereum mainnet would kill this model. $5+ gas fees for every mint, stake or loan payment. USDai needs frequent onchain interactions (minting tokens, staking, tokenizing GPUs as NFTs, processing loan payments, running auctions on defaults) Low costs + fast speeds = DePIN operators can actually access capital in minutes.
9/ The Allo Sale mechanism had two paths: ICO Alignment: 70% supply, higher multipliers, no native yield, required KYC Airdrop Alignment: 30% supply, lower multipliers, instant yields, no KYC 75M sale was filled in 52 sec at $580M cap.
The $75m USDai cap raise filled in 52 seconds with 2,631 wallets participating. From $10m whales to <$100 minters, the raise saw strong participation across every segment, showing just diverse the USDai community has become.
10/ Backed by @hiFramework @dragonfly_xyz @Delphi_Digital @yzilabs. Raised $17.4M total. Major borrowers include Lyceum, Hydra Host, Compute Labs. Strong people are betting + using their infra.
11/ The idea is to turn illiquid hardware into liquid capital. > Give DePIN operators fast funding > Give depositors stable yields from productive assets Imo @USDai_Official is addressing a genuine bottleneck in AI infra scaling. Worth watching.
6,43 тыс.
50
Содержание этой страницы предоставляется третьими сторонами. OKX не является автором цитируемых статей и не имеет на них авторских прав, если не указано иное. Материалы предоставляются исключительно в информационных целях и не отражают мнения OKX. Материалы не являются инвестиционным советом и призывом к покупке или продаже цифровых активов. Раздел использует ИИ для создания обзоров и кратких содержаний предоставленных материалов. Обратите внимание, что информация, сгенерированная ИИ, может быть неточной и непоследовательной. Для получения полной информации изучите соответствующую оригинальную статью. OKX не несет ответственности за материалы, содержащиеся на сторонних сайтах. Цифровые активы, в том числе стейблкоины и NFT, подвержены высокому риску, а их стоимость может сильно колебаться. Перед торговлей и покупкой цифровых активов оцените ваше финансовое состояние и принимайте только взвешенные решения.