Comparing Intel's oneAPI to traditional multi-threaded CPP code

Intel's oneAPI vs CPP
Students: Danny Delic
Supervisor: Pınar Tözün
Level: BSc, Semester: Spring 2020
Tags: oneAPI, ML, heterogeneous hardware, multi-threading, data-parallel execution

Modern workloads have a need for architectural diversity, as no single architecture is suitable for every problem, when optimizing for high performance. Intel’s oneAPI aims at making programming easier for data-intensive applications across heterogeneous hardware. In this project, the goal is to investigate this new technology from Intel and its effectiveness compared to multithreaded code written in CPP.