AI compute differs from gold: GPU depreciation and physical limits reshape hedging
AccBalanced · x · 2026-08-23
AI compute is increasingly viewed as a financial asset, yet its underlying hardware has physical characteristics that distinguish it from commodities like oil or gold. GPUs depreciate rapidly, with new architectures rendering older generations less valuable every few years. Unlike oil, compute capacity cannot be stored; an idle GPU-hour today vanishes rather than becoming sellable inventory. This reality shapes financing and hedging: forward markets may price future compute lower due to efficiency gains, GPU-backed loans must handle quick value loss and illiquidity, and while cash-settled derivatives mitigate price risk, they cannot solve the physical delivery constraints.
More from Venture
- Gary Marcus: Watch Out for ARR Ambiguity at Anthropic — Gary Marcus · 2026-08-23
- Harvard dropouts' AI trading startup Spectre Intelligence joins Y Combinator — ycombinator · 2026-08-23
- Open Weights vs Frontier: Just a 3-Point Gap but 1/3 the Cost — MicahBerkley · 2026-08-23
- How Does the Market Value an AI Harness? ARR Multiples Examined — le_james94 · 2026-08-23
- Indie dev shares lifetime revenue: Real estate SaaS hits $100k — martyamark · 2026-08-23
- Major software cos with captive ecosystems should sell inference directly — boymanrobshit · 2026-08-23