A user is inquiring about the feasibility and configuration of deploying a multi-GPU environment on Hugging Face Spaces. The request specifies a need for 8 GPUs, each with more than 40GB of VRAM.
The user plans to allocate 4 GPUs for a generative model and the remaining 4 for a vLLM model within the same Space. They are seeking advice on recommended configuration steps, suitable GPU hardware tiers, resource splitting best practices, and associated pricing or quota limitations.