AI GPU compute rental — RTX training fleet
Asset-backed infrastructure — pooled capital buys and operates NVIDIA RTX-class GPU training servers, leased to AI labs and startups on monthly compute contracts. Revenue from utilization and rental fees, not model equity.
Pend Investment Studio structures an SPV that procures, racks, and operates RTX GPU training machines (power, cooling, networking, remote access) and rents capacity to vetted AI labs and startups in Egypt and the region. Minimum ticket EGP 500,000. Target economics: rental income pass-through after opex — reference band illustrative until operator model is finalized. Risks: utilization gaps, power tariffs, hardware obsolescence, tenant default, and regulatory/export controls on advanced compute. Express interest for term sheet — not advice.
Physical GPU servers (RTX-class training builds), colocation or Studio-operated rack space, lease contracts with AI tenants, maintenance and refresh reserve.
Research Pend publishes — not investment advice. Users apply their own mandate.