> For the complete documentation index, see [llms.txt](https://pin3.gitbook.io/pin3-document/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://pin3.gitbook.io/pin3-document/use-cases/compute-power-contribution.md).

# Compute Power Contribution

GPU providers are rewarded based on the compute power they contribute to the PIN3 network. The more compute power a GPU provider offers, the greater their potential rewards. This encourages GPU providers to allocate a significant portion of their idle GPU compute capability to the platform, ensuring a robust and reliable supply of GPU resources for AI training tasks.