search

NVIDIA-AI-IOT / yolo_deepstream

yolo model qat and deploy with deepstream&tensorrt
Apache License 2.0
549 stars 139 forks source link

shared memory using Jetson devices #39

Closed jjzhu closed 9 months ago

jjzhu commented 1 year ago

in pushImg, we have to copy memory from cpu to gpu or device to device. for Jetson devices, such as AGX orin, can we save this step by leveraging shared memory.

wanghr323 commented 1 year ago

Yeah, good idea, Will add this feature once we have time.

wanghr323 commented 9 months ago

we won't develop this feature as this sample mainly guidance for QAT and deepstream deploy