Hartree Centre at SC18

11-16 Nov 2018 | STFC exhibited at the world’s largest HPC conference in Dallas, Texas USA.

This year the Supercomputing 2018 (SC18) conference celebrated its 30th anniversary. Bringing together over 13,000 delegates and 386 exhibitors together to share the latest developments in HPC across the globe. We were pleased to be part of the Science and Technology Facilities Council (STFC) stand alon​gside the Scientific Computing Department. It was the team’s 20th year at SC and we contributed to a variety of sessions in the technical programme as well as talks at vendor booths and interactive demos and posters of our work with industry at the stand.

To kick off our involvement in SC18, Hartree Centre Director Alison Kennedy presented a keynote at The Latest Advances in Scalable Algorithms for Large-Scale Systems workshop which was organised and chaired by Vassil Alexandrov, a Visiting Scientist at the Hartree Centre. The workshop focused on discussions around novel scalable scientific algorithms that are needed to enable key science applications to exploit the computational power of large-scale systems. Alison’s talk gave examples of the work being undertaken at the Hartree Centre, applying emerging technologies and scalable algorithms to industry challenge-led projects.

Our Training and Events Manager, Nia Alexandrov co-chaired the 5th Workshop on Best Practices for HPC Training and Education. This full-day workshop was provided opportunities for extending collaborations among practitioners from traditional and emerging fields, exploring the challenges in developing and deploying HPC training and education, and identifying new challenges and opportunities for the latest HPC platforms. Lessons learned and results from the workshop will be captured and disseminated in a Special Edition of the Journal of Computational Science Education.

James Clark, one of our High Performance Software Engineers was selected to take part in the SC18 Early Careers programme which provided a series of special sessions to help participants secure a better understanding of the issues and challenges faced while navigating a successful research career.

We were pleased to meet a range of delegates throughout the conference exhibition at the STFC booth. Our HPiC Raspberry Pi cluster was on hand to show off some demos of our work and staff were available to talk about our projects with industry. Colleagues from the Scientific Computing Department and IBM Research presented one of the top five research posters at SC18, GPU Acceleration at Scale with OpenPower Platforms in Code_Saturne.

The rest of the team took full advantage of the diverse SC18 technical programme, attending workshops for HPC systems professionals, Research Software Engineers and Women in HPC as well as tutorials on containers and collaboration and user support tools. ​

We had a great time sharing our work on an international stage and enjoyed hearing about the latest developments in HPC, data, systems, networking and storage. We can’t wait do it all again next year in Denver, Colorado. Until then, hear what some of our team had to say throughout the conference:​​​​

Great morning talk: The Human Side Of Data Science. As HPC systems get more advanced & solve large, more complex problems, we mustn’t forget about the human side. You can do all the DS you like but if no-one uses it due to inaccessibility then it is worthless. #HPCInspires#SC18

— Tim Powell (@MrTimPowell) November 15, 2018

Attending #IXPUG BoF at Supercomputing sharing experiences about optimizing for Intel processors #IntelHPC @HartreeCentre pic.twitter.com/K9XYg7VReH

— Sergi Sisó (@sergiesg) November 15, 2018

We must be scrupulous, have our results be reproducible, and have our data sources publicly available in order to be robust to accusations of ‘fake news’ and be able to produce reliable science. #HPCInspires #SC18

— Tim Powell (@MrTimPowell) November 13, 2018