19208

Benchmarking Deep Learning Models on Jetson TX2

Lucas Pedro Bordignon, Aldo von Wangenheim
INCoD – Brazilian Institute for Digital Convergence
Federal University of Santa Catarina, 2019

@article{bordignon2019benchmarking,

   title={Benchmarking Deep Learning Models on Jetson TX2},

   author={Bordignon, Lucas Pedro and von Wangenheim, Aldo},

   year={2019}

}

In conclusion, the present work brings an overview of artificial intelligence and, mainly, deep learning fields with a focus on image recognition and the history behind the models and techniques present nowadays. Beyond that, we explored how embedded hardware work with the new scenarios that AI brings to the table and how companies are developing task-specific boards to tackle new problems. After that, we worked with one of such hardware, called NVIDIA Jetson TX2, explaining how to flash, setup and use it with a hands-on tutorial. Finally, we developed a benchmark to measure how well the Jetson platform behaves under real-world circumstances and explore it results, bringing more information to the decision-making process when deciding which architecture better fits the present hardware. As future work, the application of the TensorRT framework on the benchmark, developed by the manufacturer, is expected and the execution of experiments under different datasets and with new models to enrich the data presented by the current work.
No votes yet.
Please wait...

* * *

* * *

HGPU group © 2010-2019 hgpu.org

All rights belong to the respective authors

Contact us: