We have chutes.ai for distributed GPU computing and it seems to be quite successful.

https://chutes.ai/platform

Tokens processed through OpenRouter gateway: https://openrouter.ai/provider/chutes