Hi,
I want to install ChatRTX on a server with Nvidia A100 GPU on it. Even if A100 has Ampere architecture, the installation fails at “system check” step. The error is as follows:
NVIDIA Installer cannot continue
“ChatRTX” is a supported on Ampere and above GPU family
“ChatRTX” requires minimum 8 GB of GPU memory to function.
Considering A100 has 40 GB memory and its architecture is Ampere, shouldn’t it work?
In ChatRTX Github page (GitHub - NVIDIA/ChatRTX: A developer reference project for creating Retrieval Augmented Generation (RAG) chatbots on Windows using TensorRT-LLM) it says:
“ChatRTX is currently built for RTX 3xxx and RTX 4xxx series GPUs that have at least 8GB of GPU memory.”
Does this mean ChatRTX cannot run on A100 GPU?
If anyone has been able to run ChatRTX on the A100 GPU before or knows a way to do this, I’d be very grateful if they could help me out.
Best regards