-------------------------------------------------------------------------------------------------------
Initial try.  Only H264.

https://developer.nvidia.com/blog/nvidia-ffmpeg-transcoding-guide/

# example, H265 recoding of a ~1GB '.ts' stream of duration ~25min

/opt/ffmpeg/6x-cuda/bin/ffmpeg  -hwaccel cuda -hwaccel_output_format cuda -i ~/tst/in.ts -c:v h264_nvenc -b:v 1.2M /tmp/oot.mp4

# runs a fair few times faster than on the cpus:  e.g. 16x play, instead of ~0.4 play-speed
# but haven't found a controllability like the tune and preset and quality settings of the cpu version 
# *and* it doesn't appear to do H265 encoding with the nvidia thing (what's 'hevc_vulkan'?) 

end of ffmpeg :
video:210962kB audio:23647kB subtitle:0kB other streams:0kB global headers:0kB muxing overhead: 0.772789%
frame=75002 fps=848 q=21.0 Lsize=  236422kB time=00:25:00.01 bitrate=1291.2kbits/s speed=16.9x    
[aac @ 0x428d3c0] Qavg: 674.545
real    1m34.365s

during this time, not very high loading:

$ nvidia-smi 
| 
Mon Jun 12 21:42:34 2023
+---------------------------------------------------------------------------------------+
| NVIDIA-SMI 530.30.02              Driver Version: 530.30.02    CUDA Version: 12.1     |
|-----------------------------------------+----------------------+----------------------+
| 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 A40                      Off| 00000000:41:00.0 Off |                    0 |
|  0%   32C    P0               85W / 300W|    377MiB / 46068MiB |      3%      Default |
|                                         |                      |                  N/A |
+-----------------------------------------+----------------------+----------------------+

+---------------------------------------------------------------------------------------+
| Processes:                                                                            |
|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |
|        ID   ID                                                             Usage      |
|=======================================================================================|
|    0   N/A  N/A   3812595      C   /opt/ffmpeg/6x-cuda/bin/ffmpeg              373MiB |
+---------------------------------------------------------------------------------------+


-------------------------------------------------------------------------------------------------------
More. 

Wanting H265 (HEVC). 
Can't (briefly) find any option (actually effectual) like the '-crf' for libx265, to give quality-based 
variable rate.  

The incantation was:
 ffmpeg  -hwaccel cuda -hwaccel_output_format cuda -i in.ts  -c:v hevc_nvenc -b:v 1.5M cuda_hevc1M5.mkv 
this does pretty fast (~15x on the A40 on 

See in particular this page (or corresponding pdf: "Using FFmpeg with NVIDIA GPU Hardware Acceleration")
  https://docs.nvidia.com/video-technologies/video-codec-sdk/12.0/ffmpeg-with-nvidia-gpu/index.html