Back to Newsroom

Nvidia's new $500B plan is risky but brilliant, especially for aging GPUs

By Modelverse Editorial·August 13, 2026·2 min read
Nvidia's new $500B plan is risky but brilliant, especially for aging GPUs

Nvidia's recent announcement to commit up to $500 billion to build AI data centers has sparked interest in the company's effort to create a secondary market for aging GPUs. The plan involves Nvidia guaranteeing the value of its chips used as collateral in deals, with the company covering up to 25% of the difference if the GPUs don't retain their expected value. This move aims to sustain demand for Nvidia hardware as it ages and ensure an ecosystem of used AI hardware flourishes.

Key aspects of Nvidia's plan include:

  • Guaranteeing the value of chips used as collateral
  • Covering up to 25% of the difference if GPUs don't retain their expected value
  • Creating a secondary market for aging GPUs to sustain demand for Nvidia hardware

The plan poses a "wrong way" risk for Nvidia, where its obligations grow as demand weakens, potentially squeezing revenues. Despite this risk, Nvidia's scheme differs from comparisons to Lucent Technologies, with the company deliberately structuring its commitments to mitigate potential downsides. Nvidia's efforts to create a thriving ecosystem for used AI hardware may have significant implications for startups and enterprises, particularly in terms of hardware licensing and utilization.

ai-newsbreakingtechcrunch-ai

Footnotes & Primary References

Related content

Writer introduces new AI model and upgraded harness to contain token costs

Built as a post-training variation on Z.ai's open source model GLM-5.2, Writer says the new system should provide deployment-ready capabilities at a much lower price.

Read article

OpenAI introduces 'Ultrafast,' a new mode that makes GPT-5.6 Sol work at 14x the speed

OpenAI is launching a preview of a sped up version of its latest, most powerful model, in an effort to court enterprise users.

Read article

Xiaomi's MiLM Plus Releases PROVE: Perception-Aligned Object Removal Metrics RC-S and RC-T With a Real-World Video Benchmark

Object removal models have improved faster than the metrics used to judge them. Diffusion erasers now reconstruct shadows, reflections and occluded structure convincingly, yet PSNR...

Read article