Accelerating Open Source Science Applications with NVIDIA Tesla V100
Summary
In January 2019, Isabella computer cluster in University computing centre in Zagreb (Srce) was expanded with the latest generation of graphics processors - NVIDIA Tesla V100 based on Volta architecture, designed for high performance computing, AI and machine learning workloads. Graphics processing units are a part of most modern supercomputers and computer clusters because they enable acceleration of user applications by multiple orders of magnitude. Integrating new GPUs in Isabella cluster involved adapting open source scientific applications and machine learning frameworks for execution on graphics processors, as well as developing job scheduling mechanisms to efficiently use V100's compute capabilities in a shared user environment.
Type
Talk
Language
English
Speaker
Tomislav Smolčić works in the University computing centre (SRCE) as HPC system administrator. He is part of a team dedicated to advancement of e-science and use of new technologies and resources in different fields of science and industry.
