Thoughts on using a RTX2070 Super with a mini PCIe adapter?
I am in needing of a somewhat portable set up to run some deep learning applications.
I was thinking in using an old laptop with a Core i7 2670QM CPU and 8 GB of RAM connected to an RTX 2070 Super through a mini PCIe port (currently being used by a WiFi card).
However, I was looking the forum and haven't found anyone using RTX cards with mini PCIe adapters.
So, I am not sure if it even works. Also, I don't know if my mini PCIe port is whitelisted or not. The laptop is from a Brazilian brand (CCE) from 2011, which was basically created by a chinese company called Topstar (TPS). Hence, I did not found anything online about eGPUs with this laptop.
I know the card would be connected to only a single lane, and it would be probably be pretty slow. However, I have no intention in using this set up to play games. I only want to use PyTorch with CUDA and Linux. I don't even want to connect the card into a monitor. I am also thinking in going with an RTX model because it has tensor cores which are faster in FP16 than simple cuda cores.
1) Would an RTX 2070 Super worked using a mini PCIe port?
2) Is it possible to verify if my mini PCIe port is whitelisted without an eGPU?
Thanks in advance for any suggestion or information you may have.
Is it possible to verify if my mini PCIe port is whitelisted without an eGPU?
Try connecting a cheap USB PCIe card?