You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Question
Looking for opinions here on ideas why the 3080 was faster. I am running a YoloV5s model exported to Onnx and running via OpenCV and C++ with CUDA 11.7. I was doing some baselining this weekend and noticed my 3080 was running the network faster than a 4090. Same desktop PC, only swapping the GPU, then running inference on the exact same dataset of 5000 images.
3080 got 4.8ms for a net.forward() pass whereas the 4090 was at 6.8ms.
Details
CPU AMD 7950X, DDR5-6400, NVME PCI-E 4 SSDs.
Images were 640x480 resized to 640x640.
Anyone know why or have similar experiences? My thoughts are that the software may be more optimized for the 3080?
I found another question that was similar, but overall unanswered: #10667
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Question
Looking for opinions here on ideas why the 3080 was faster. I am running a YoloV5s model exported to Onnx and running via OpenCV and C++ with CUDA 11.7. I was doing some baselining this weekend and noticed my 3080 was running the network faster than a 4090. Same desktop PC, only swapping the GPU, then running inference on the exact same dataset of 5000 images.
3080 got 4.8ms for a net.forward() pass whereas the 4090 was at 6.8ms.
Details
CPU AMD 7950X, DDR5-6400, NVME PCI-E 4 SSDs.
Images were 640x480 resized to 640x640.
Anyone know why or have similar experiences? My thoughts are that the software may be more optimized for the 3080?
I found another question that was similar, but overall unanswered: #10667
Beta Was this translation helpful? Give feedback.
All reactions