From: 2020-05-12 10:15 to: 11:00
Place: https://lu-se.zoom.us/j/63826191255
Contact: birger [dot] swahn [at] cs [dot] lth [dot] se
Save event to your calendar

One Computer Science MSc thesis to be presented on 12 May 2020 via Zoom

Tuesday, 12 May there will be a master thesis presentation in Computer Science at Lund University, Faculty of Engineering.

The presentation will take place via Zoom at: https://lu-se.zoom.us/j/63826191255

Note to potential opponents: Register as an opponent to the presentation of your choice by sending an email to the examiner for that presentation (firstname [dot] lastname [at] cs [dot] lth [dot] se). Do not forget to specify the presentation you register for! Note that the number of opponents may be limited (often to two), so you might be forced to choose another presentation if you register too late. Registrations are individual, just as the oppositions are! More instructions are found on this page.

10.15-11.00

Presenter: Martin Karp
Title: Optimization and Profiling of the Spectral Element Method on Graphics Processing Units
Examiner: Krzysztof Kuchcinski
Supervisors: Jonas Skeppstedt (LTH), Niclas Jansson (KTH)

In this work, optimizations of the Spectral Element Solver Nek5000 on single and multiple GPUs were investigated. In particular, in the proxy app Nekbone, the main computation was optimized for a single GPU with a mixed OpenACC and CUDA approach. In addition, measurements were taken to asses the scaling of the code for multiple GPUs. The results showcase that a close to optimal single GPU implementation of Nekbone had been achieved. For future performance increases however, it was concluded that the current communication kernel is the main bottleneck.