RPO: The Currency of Neocloud Hype

Summary

• RPO (Remaining Performance Obligation) has become the new metric and currency of the circular investment in AI and AI infrastructure.
• The quality of a neocloud’s RPO pipeline is important to evaluate, especially as circular investment deals/arrangements continue to get larger and more risky/creative.
• Token economics will inevitably put neocloud RPO pipelines at risk as token pricing continues to drop, and AI infrastructure systems and technologies face imminent performance and economic scaling walls in the coming quarters and years.

neXt Curve Insights

You have probably heard about RPO or Remaining Performance Obligation from the neoclouds and big hyperscalers who are currently investing hundreds of billions into the build-out of gigawatt-scale AI data centers and infrastructure around the world.

RPO was introduced to GAAP accounting with the adoption of the new revenue recognition standard, ASC 606, which became effective for public companies in 2018 and for private companies starting in 2019. This new standard made RPO a mandatory disclosure for public companies to improve financial transparency for investors.

RPO is widely reported by SaaS companies who are engaged in long-term contracts with their customers. RPO reporting provides a pipeline view into the future revenue of existing contracts and purchase orders, which can include deferred revenue and unbilled revenue.

Over the past couple of years, neoclouds and cloud service providers with growing neocloud businesses have been floating massive RPO numbers and growth for their GPUaaS (GPU as a Service) businesses. Most notably, Oracle’s reporting of RPO of $455 billion in the first quarter of 2025 (fiscal Q1 2026), a 359% increase YoY, raised eyebrows. Oracle’s massive RPO increase in the quarter was largely attributed to their Stargate-related deal with OpenAI in the order of $300 billion.

Unlike RPO of SaaS businesses, the unrealized revenue of neocloud RPO pipelines are anchored to titanic AI infrastructure investments and supercomputing capacity build-out. Once these buildouts are completed, neoclouds are subject to depreciation expense of AI systems subject to a one-year cadence and aggressive scaling dynamics, a growing concern among industry observers and investors.

In large part, GPaaS RPO’s are turning out to be based on massive commitments of varying quality and structure between neocloud players and a small number of unprofitable AI service providers (A.K.A. AI labs), such as OpenAI and Anthropic.

This has raised alarms in the industry as circular investment arrangements continue to make headlines on seemingly a weekly basis.

(Click here for the full article)

This material may not be copied, reproduced, or modified in whole or in part for any purpose except with express written permission or license from an authorized representative of neXt Curve. In addition to such written permission or license to copy, reproduce, or modify this document in whole or part, an acknowledgement of the authors of the document and all applicable portions of the copyright notice must be clearly referenced.

© 2025 neXt Curve. All Rights Reserved