What is the appropriate CUDA TK and cuDNN version for the following GPU card and driver?

What is the appropriate CUDA TK and cuDNN version for the following GPU card and driver?

# GeForce GTX 780
# Kepler architecture
#'
#+-----------------------------------------------------------------------------+
#| NVIDIA-SMI 470.223.02   Driver Version: 470.223.02   CUDA Version:          |
#|-------------------------------+----------------------+----------------------+
#| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
#| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
#|                               |                      |               MIG M. |
#|===============================+======================+======================|
#|   0  NVIDIA GeForce ...  Off  | 00000000:03:00.0 N/A |                  N/A |
#| 33%   30C    P8    N/A /  N/A |     14MiB /  3020MiB |     N/A      Default |
#|                               |                      |                  N/A |
#+-------------------------------+----------------------+----------------------+
#|   1  NVIDIA GeForce ...  Off  | 00000000:84:00.0 N/A |                  N/A |
#| 33%   30C    P8    N/A /  N/A |      6MiB /  3022MiB |     N/A      Default |
#|                               |                      |                  N/A |
#+-------------------------------+----------------------+----------------------+
#
#+-----------------------------------------------------------------------------+
#| Processes:                                                                  |
#|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
#|        ID   ID                                                   Usage      |
#|=============================================================================|
#|  No running processes found                                                 |
#+-----------------------------------------------------------------------------+
#'

Techpowerup is a good place to find card information, in this case, the Compute Capability, (or “Cuda” in the “Graphics Features” box), is 3.5.

Knowing this, the range of Cuda Toolkits supporting it, can be found at the Wikipedia page for Cuda, in the first table of the section “GPUs Supported”, so Toolkit 5.0 - 11.8.

cuDNN for older GPU’s, is a case of going to the documentation archive and digging through the Support Matrix for each version.