ACADEMIC PAPER SUMMARY.

 

                                   (FRONT PAGE OF THE ACADEMIC PAPER)

The paper title is "Structure of  Deep learning Inference Engines for Embedded systems". This paper talks about deep learning incorporated with embedded systems, it first talks about how deep learning has been used to solve some unsolvable problems it then proceeds to talk about how deep learning applications require computer accelerators like GPUs, this is due to the enormous data and technical operations, also it stated the downside of embedded systems in the environment, causing an increase in temperature and requires a lot of power to function.

The main focus point of deep learning with embedded systems in this paper is the application with the development of automobile engines, the bases of design used were PlaidML and the deep interference engine A.HPC. PlaidML is an open-source-based deep learning framework developed by Intel. It supports a user interface and it is convenient, it does not run on a lot of GPUs whereas, the A.HPC runs on a lot of GPUs and the physical size is limited, also they are power-consuming and increase in temperature.

This article is available at : IEEE Xplore Full-Text PDF: