ohjoij-ys

Results 10 comments of ohjoij-ys

Multi GPU not work in my PC. system:ubuntu 20.04 gpu : rtx3080 x2 drivers: 560 (not open) ![image](https://github.com/user-attachments/assets/d34ddf88-4568-42c3-9238-fd27936daeca) 1gpu ![image](https://github.com/user-attachments/assets/d645edf2-3886-4a71-a7f1-42e002500113) 2 gpu ![image](https://github.com/user-attachments/assets/0759e10f-24d6-4edd-8d17-c63cc1c610b1) ![image](https://github.com/user-attachments/assets/1a467859-cf20-4347-8f4e-19fa73fe273e)

and same result in the windows ,I'm pretty sure it has identified all the cards

use ZoeD_N ALLcuda - 3.8 FPS Singal GPU -5.3 fps ![image](https://github.com/user-attachments/assets/ab9804a0-1b23-4771-8af9-8ffcd88811f5) ![image](https://github.com/user-attachments/assets/bd14a9ee-1359-471f-85cd-b9b7a1c89a7c) ![image](https://github.com/user-attachments/assets/c1399398-d3a6-4be0-9bf9-f17577dade40) ![image](https://github.com/user-attachments/assets/8bfb8a57-9695-4733-8da2-53da81e5adaf) IF I Manually specifying the number of threads to call will result in out of memory...

use Any_V2_N_S (Doesn't seem to support multi gpu) All_cuda : 4.85 Fps singal gpu: 5.0 Fps ![image](https://github.com/user-attachments/assets/432ce251-faf0-4ba9-aed9-f73285e4b8d7) ![image](https://github.com/user-attachments/assets/5e42b0bc-d4b9-4a87-b1d5-7c8f69becfc5) ![image](https://github.com/user-attachments/assets/e078e3ac-837f-46da-aa16-059f10156ccd) ![image](https://github.com/user-attachments/assets/9577ffa2-f891-463c-9208-ccfdb809660a) ![image](https://github.com/user-attachments/assets/377b02bf-01fb-464c-a7dd-b6b6d29d3d25)

all_cuda vs singal gpu it seems all not support ,driver/cuda version problem? any_s : 4.91 vs 4.91 Any_B : 4.81 vs 4.81 Any_V2_N_B :4.77 vs 4.85 zoeD_K 4.06 vs 4.2

it's not work ![image](https://github.com/user-attachments/assets/187c6c31-dcc1-47cf-b5e2-e5229c32fc27)

max batch size is 4 (it will be out of vram when higher) multi gpu 3.84 fps singal gpu 5.30 fps it's reduce performance? ![image](https://github.com/user-attachments/assets/6e6e2504-0482-4d1c-99f5-fbae19545a0f) ![image](https://github.com/user-attachments/assets/bdc43842-ff5c-41a0-a41e-43fc68c88896) ![image](https://github.com/user-attachments/assets/ebeafa30-5958-42ef-8680-bf6d9438ca12) ![image](https://github.com/user-attachments/assets/f8b5c1d7-be75-4ec8-a589-44db8294ed79)

batch-size = 8 for gpu ![image](https://github.com/user-attachments/assets/19066601-4dc9-4968-9ea3-60024c70365c) multi gpu 4.2 Fps ![image](https://github.com/user-attachments/assets/a7d31785-2e73-423c-b594-852a5eac5d0b) singal gpu 5.9 Fps ![image](https://github.com/user-attachments/assets/ca751b9c-8574-4d09-b1bd-fad2d351a16a)

ANY_B low-resolution video bath_size = 16 singal gpu,fps :38 multi-gpu ,fps:38 ![image](https://github.com/user-attachments/assets/6e911b61-f9b8-45a4-9e6f-62575c4c3d96) ![image](https://github.com/user-attachments/assets/50bc4a8b-1fd0-488d-9085-a57e98f278ed) bath_size =32 multi gpu fps:38 singal gpu fps:38 cpu load mode seems be different ![image](https://github.com/user-attachments/assets/3b915d76-58a9-4281-ad36-9ae7594958b0) ![image](https://github.com/user-attachments/assets/54279957-27fc-4998-805e-1860e5fbeda5) ![image](https://github.com/user-attachments/assets/27997fd7-782a-468d-b048-cbca4c5e9661)...

Maybe it’s because Nvidia has cut some features from gaming graphics cards compared to professional cards. Anyway, I’m looking forward to your new multi-GPU strategy.