[W] [TRT] Unknown embedded device detected. Using 59656MiB as the[W] [TRT] Unknown embedded device detected. Using 59656MiB as the allocation ca

Robotics & Edge Computing Jetson & Embedded Systems Jetson AGX Orin
1

When i use: trtexec --onnx=yolov5.onnx --saveEngine=yolov5.engine
I met this problem:
[W] [TRT] Unknown embedded device detected. Using 59656MiB as the allocation cap for memory on embedd
tegrastats shows:
8:25:49 RAM 5033/62797MB (lfb 13875x4MB) SWAP 0/31398MB (cached 0MB) CPU [16%@1728,36%@1728,22%@1728,27%@1728,23%@960,9%@1344,21%@1344,10%@1344,off,off,off,off] EMC_FREQ 19%@2133 GR3D_FREQ 99%@611 GR3D2_FREQ 99%@0 VIC_FREQ 115 APE 174 CV0@-256C CPU@50.218C Tboard@38C SOC2@46.218C Tdiode@39.25C SOC0@46.906C CV1@-256C GPU@46.343C tj@50.218C SOC1@46.218C CV2@-256C VDD_GPU_SOC 6394mW/6394mW VDD_CPU_CV 1599mW/1599mW VIN_SYS_5V0 6471mW/6471mW NC 0mW/0mW VDDQ_VDD2_1V8AO 2019mW/2019mW NC 0mW/0mW

my device is AGX orin 64G, ubuntu20.04,TensorRT v8502,cuda 11.4,jetack 5.1

2

.[W] [TRT] Unknown embedded device detected. Using 59656MiB as the[W] [TRT] Unknown embedded device detected. Using 59656MiB as the allocation ca

4

Hi,

This is a known issue for Orin 64GB board and is fixed in the upcoming release.
The message is a harmless warning and TensorRT still can work correctly.

Thanks.

6

TensorRT 8.4.1.5 seems to be way faster than version 8.5.2.2 that prints me the same warnings.

7

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.