Banca de QUALIFICAÇÃO: LEONANDRO VALERIO BARBOSA GURGEL

Uma banca de QUALIFICAÇÃO de MESTRADO foi cadastrada pelo programa.
STUDENT : LEONANDRO VALERIO BARBOSA GURGEL
DATE: 11/07/2024
TIME: 08:00
LOCAL: Sala Google Meet - https://meet.google.com/aja-yvyp-qki?hs=224
TITLE:

Utilizing Osmotic Computing for Distribution of Deep Learning Models

KEY WORDS:

Edge Computing, Cloud Computing, Distributed Deep Learning, Osmotic Computing.

PAGES: 49
BIG AREA: Ciências Exatas e da Terra
AREA: Ciência da Computação
SUBÁREA: Metodologia e Técnicas da Computação
SPECIALTY: Engenharia de Software
SUMMARY:

Solutions that use edge computing have benefited from the low latencies
from processing on servers in close proximity or even on the edge device.
device, while solutions that use cloud computing take advantage of the great
centralized computing power. One prominent area that utilizes both cloud
computing is applications that use deep learning models.
learning models. In this sense, even though edge computing can provide very low response
response time, the devices usually have very limited computational resources.
computational resources, while cloud computing provides a great deal of computational power but
end up suffering from high access latencies. As many of these applications
applications that use deep learning cannot tolerate one of these restrictions.
scalability when choosing to focus on Edge or Cloud or even choosing a fixed
a fixed distribution model. In this context, osmotic computing has emerged as an
both the cloud layer and the edge layer, making better use of resources.
better use of resources. With this in mind, this paper proposes a distribution model for
models using osmotic computing through the SAPPARCHI platform.
SAPPARCHI platform. This work: (i) presents a solution that uses a deep learning model
learning model, (ii) proposes a solution using SAPPARCHI and another using a microservice-based
model based on microservices, (iii) evaluates the performance and scalability of the two solutions.

COMMITTEE MEMBERS:
Presidente - 1678918 - NELIO ALESSANDRO AZEVEDO CACHO
Interno - 2316877 - EVERTON RANIELLY DE SOUSA CAVALCANTE
Externo ao Programa - 2669476 - ARTHUR EMANOEL CASSIO DA SILVA E SOUZA - UFRNExterno ao Programa - 1669545 - DANIEL SABINO AMORIM DE ARAUJO - UFRNExterno ao Programa - 2510306 - FREDERICO ARAUJO DA SILVA LOPES - UFRN