Embedded cluster platform for a remote parallel programming lab

An Open edX Extension for Parallel Programming Assignments with Automatic Configurable Grading

“…This paper is a continuation of our preliminary results published in [19], where the platform was presented with less detail and the focus was on the study performed to determine the best SBC to build the cluster. In this work, we present the whole platform in detail, discussing the strengths and limitations of the most relevant components.…”

“…We have chosen such devices, instead of traditional desktop-or server-like clusters, given their advantages in terms of low cost, low power consumption and recent utilization for edge computing. In fact, we acquired seven different cards (Vim3Pro, Orange Pi4B, Odroid N2, Hikey 970, Jetson Nano, Raspberry Pi4B, and Latte Panda) and evaluated them in terms of processing latency (by running standard benchmarks), price, and software compatibility [19]. Such devices are based on ARM processors typically found on embedded electronic devices and cellphones.…”

Incentivando a adoção de modelos de programação concorrente/paralela a partir da visualização de ganhos na geração de fractais em tempo-real

“…Graças ao advento das arquiteturas multicore, somado a maior acessibilidade a recursos de processamento, a programação concorrente/paralela assumiu um papel de destaque na indústria de software [Wilhelm et al 2018]. Portanto, programadores devem possuir habilidades neste contexto, pois passam a ser fundamentais para explorar toda a potencialidade dos novos sistemas computacionais [Sarkar et al 2018, Velásquez et al 2020. Ainda assim, o ensino tradicional, nos cursos de graduação na área da computação, consideram este tópico como avançado, muitas vezes considerado como eletivo [Giacaman and Sinnen 2018].…”