ocf_still_01.jpg

Bede Training Resources and Support

Bede is the N8’s newest high-performance computing platform. It has a GPU-accelerated architecture that makes it ideal for artificial intelligence and machine learning applications.

Bede was originally handed over in August 2019 and since then has been commissioned and tested before being opened to researchers. More than 50 projects are currently running on Bede, the majority of these are from within the N8 Research Partnership, but several are researchers accessing the system via HECBioSim.

You can find out how to apply for time via the website: https://n8cir.org.uk/supporting-research/facilities/bede/docs/bede_registrations/

Because Bede uses IBM Power 9 CPUs rather than a more traditional x86 architecture, each of the N8 universities has a full-time equivalent research software engineer (RSE) dedicated to supporting researchers at every stage of the process. They can offer advice about the amount of memory and storage you may need as well as advising on code and many different aspects of integrating high-performance computing (HPC).

You can find out more about your local RSE support here: https://n8cir.org.uk/supporting-research/facilities/bede/rse-support-bede/

Alongside the site-specific support, there has also been extensive training and support sessions from IBM. These include:

  • Hardware and Software
  • GPU Programming
  • Application Porting and Tuning
  • AI Workloads
  • Watson Machine Learning Community Edition
  • Large Model Support

You can find all of these videos at: https://n8cir.org.uk/events/event-resource/ibm-training-bede/

At an introductory event in January 2021 a number of researchers presented case studies of their existing or planned work on Bede. These short presentations can be seen at: https://n8cir.org.uk/events/event-resource/bede-intro-21/

Return to article index