High-Level Overview
Compute Labs is a Pasadena-based technology startup that tokenizes enterprise GPUs into yield-bearing digital assets like GNFTs and COMPUTE tokens via its Compute Tokenization Protocol (CTP), enabling investors to gain direct exposure to AI compute infrastructure while generating 20-50% yields from live AI workloads.[1][2][4] It serves accredited investors, AI companies, cloud providers, and HPC operators by purchasing GPUs on behalf of investors, leasing them to data centers on a revenue-sharing model, and handling deployment, management, and compliance, thus solving GPU supply shortages and high capital costs in the AI boom.[2][4] The company has demonstrated strong early growth, launching a $1M NVIDIA H200-backed GPU vault in June/July 2025 that fully deployed within weeks, raising $3M in pre-seed funding led by Protocol Labs, and lining up over $100M in GPU pipeline post-pilot.[2][4]
Origin Story
Founded in March 2024 by Albert Zhang, Compute Labs emerged from his vision to create a GPU Real World Asset (RWA) project amid surging AI compute demands.[1][3] Zhang, previously a founding team member at Delysium (a $2B peak FDV AI-agent network), core member of rct.AI (YC19), and Solution Architect at Xsolla, holds degrees from UCLA and Caltech; he leads as CEO alongside CTO Xingfan Xia (ex-Apple, AWS, Airbnb), CBO Nikolay Filichkin, and founding engineers like Marcelo Feitoza and Matt Weichel.[1][3] Incubated by NVIDIA Inception VC Alliance, the company quickly gained traction with its $3M pre-seed round and pilot vault launch, addressing crypto-AI finance gaps through blockchain tokenization.[2][3][4]
Core Differentiators
- Tokenization Protocol (CTP): Converts physical GPUs into verifiable, liquid digital assets (GNFT & COMPUTE tokens) on blockchain, enabling trading on exchanges and transparent ownership.[1][2]
- Yield-Generating Vaults: Investors earn high yields (20-50%) from revenue-shared leases to data centers running real AI workloads, with Compute Labs managing all operations and compliance.[2][4]
- AI-Fi Ecosystem Expansion: Offers compute derivatives like GPU ETFs, indexes, perps, staking/lending, and PT/YT to democratize AI infrastructure access beyond accredited investors.[1][2]
- Security and Open-Source Focus: Employs formal verification, bug bounties, and open-sourced smart contracts/GPU tools to ensure robustness and foster community contributions.[1]
- Cost Efficiency for AI Firms: Allows AI companies to launch custom GPU RWA vaults, sharing compute revenue to prioritize R&D over infrastructure.[1]
Role in the Broader Tech Landscape
Compute Labs rides the AI infrastructure boom, where global GPU shortages (100% utilization) drive high yields, fueled by large language models and data center demands.[4] Its timing aligns with regulatory tailwinds like the Genius Act, clarifying stablecoins and RWAs, while blockchain enables liquidity in a traditionally illiquid asset class.[4] Market forces favoring it include hyperscaler CapEx constraints and the need for scalable financing, positioning Compute Labs as a "backbone for AI" by securing GPU supply for projects and providers.[2] It influences the ecosystem by financializing compute as an emerging asset class, blending AI, blockchain, and DeFi to broaden participation and accelerate innovation.[1][2][3]
Quick Take & Future Outlook
Compute Labs is poised to scale rapidly with $10M+ raises planned, a $100M+ GPU pipeline, and potential SPAC merger for public market access, capitalizing on unrelenting AI compute hunger.[4] Trends like AGI pursuits, RWA tokenization growth, and DeFi derivatives will propel its AI-Fi ecosystem, potentially evolving it into a core AI infrastructure financier. As GPU scarcity persists, its model could redefine investing in tech hardware, tying back to its origins in tokenizing the "picks and shovels" of the AI gold rush.