1x NVIDIA Tesla V100 32GB SXM2 GPU w/ PCIe Adapter for AI Computing.
It is a fun card to have.
A PCIe version of V100 server GPU in a third-party case with built-in fan, in good condition, working as it should.
The fan is loud in full power, if reduce GPU power and limit its frequency, and add a speed control unit for the fan, it can work quietly even when thinking hard without much compromising speed.
It does work for ComfyUI on selected workflows using GGUF models, decent generating speed, it doesnt support BF16 or FP8 models. It doesn't have video output sockets.
Do not buy if you intend to use ComfyUI only, it supports enough formats for now, but it is ineed dated, you might regret not getting a 3090 or a6000 in the future. If you are looking for something cheap to speedily run LLM, this one offers best value.
Selling because no longer needed.