SemiAnalysis reports that 1-year H100 rental contract prices rose from $1.70/hr/GPU in October 2025 to $2.35/hr/GPU by March 2026 (~40%), with on-demand and spot capacity sold out across all GPU types; all new capacity coming online through August–September 2026 is already booked. The tightening is driven by a confluence of forces: surging agentic/multi-agent workloads (notably Claude Code), native media generation, open-weight model adoption, and AI-lab capital raises, while supply is simultaneously constrained by server repricing triggered by parabolic DRAM/NAND cost increases. For financial institutions, this data point materially challenges the prevailing "inevitable oversupply/commoditization" narrative embedded in depressed Neocloud valuations and raises questions about whether announced capex will actually translate into deployed capacity on projected timelines.
581 other tracked developments touching the same claims.