Environment

Software versions used across benchmark devices.

RTX 5090 (antares)

Component	Version
OS	Ubuntu 24.04 (kernel 6.8.0-100-generic)
GPU	NVIDIA GeForce RTX 5090 (32 GB VRAM)
GPU Driver	570.211.01
CUDA	12.8
cuDNN	90701
Python	3.12
PyTorch	2.7.0+cu128
Ultralytics	8.4.26
TensorRT	N/A (not used on this device)

Jetson Orin AGX

Component	Version
OS	Ubuntu (kernel TBD)
JetPack	TBD
GPU	Integrated Ampere (64 GB shared RAM)
CUDA	TBD
cuDNN	TBD
TensorRT	TBD
Python	TBD
PyTorch	TBD
Ultralytics	TBD

Jetson Orin Nano

Component	Version
OS	Ubuntu (kernel TBD)
JetPack	TBD
GPU	Integrated Ampere (8 GB shared RAM)
CUDA	TBD
cuDNN	TBD
TensorRT	TBD
Python	TBD
PyTorch	TBD
Ultralytics	TBD

How to fill in TBD values

Run the following on each device:

# OS kernel
uname -r

# CUDA
nvcc --version

# cuDNN
python -c "import torch; print(torch.backends.cudnn.version())"

# TensorRT (Jetsons only)
python -c "import tensorrt; print(tensorrt.__version__)"

# JetPack (Jetsons only)
cat /etc/nv_tegra_release
# or
sudo apt-cache show nvidia-jetpack | grep Version

# PyTorch
python -c "import torch; print(torch.__version__)"

# Ultralytics
python -c "import ultralytics; print(ultralytics.__version__)"

# GPU Driver
nvidia-smi | head -3

Development Machine (Windows)

Component	Version
OS	Windows 11 Pro for Workstations 10.0.26200
Python	3.12

This machine is used for development only, not for benchmark execution.