Every November, around 10,000 computing professionals gather in the US for The International Conference for High Performance Computing, Networking, Storage and Analysis — known in the industry simply as “SC” or colloquially as “The Zoo.”
The conference attracts researchers, system architects, system operators, developers, strategists and vendors from all over the world to a week packed with refereed workshops, tutorials, panels, birds of a feather (BoFs), vendor briefings and a fiercely competitive technical program. Even as I sat on the plane flying from San Francisco to Salt Lake — the home of SC16 — drafting this blog entry, my neighbor was working on a presentation detailing micro-optimizations to a genomics sequencing algorithm; I think I know which conference he is heading to!
SC is not exclusively about bare-metal, big-iron, batch systems (cloud computing and virtualization have featured in previous SC programs), but this year OpenStack is front and center. Following our inaugural meeting at the Austin summit, HPC practitioners in the OpenStack Scientific Working Group collaborated to submit successful proposals for panel and birds of a feather sessions focusing on the application of OpenStack for HPC.
Another noteworthy achievement of the Scientific Working Group (with a good deal of assistance from the OpenStack Foundation) during the Newton cycle was a series of research computing whitepapers pulled together by Stig Telfer. These are available as a book just in time for SC16. So, if you are interested in the integration of HPC and cloud computing, please visit openstack.org/science and participate in these sessions within the main conference program:
- 1330-1500, Tuesday, November 15 – BoF – OpenStack for HPC: Best Practices for Optimizing Software-Defined Infrastructure
- Lightning talks
- The OpenStack HPC book will be available at the BoF too
- 1530-1700, Thursday, November 17 – Panel – HPC/Research Computing: Leveraging the Architectures, Flexibilities, and Tools Emerging from the Members of the OpenStack Scientific Community
There are also a number of other highly relevant sessions of interest:
- BoF – Experimental Infrastructure and Methodology for HPC Cloud Research
- BoF – HPC Virtualization and Cloud in the OpenHPC Context
- Panel – HPC File Systems, Challenges, and Use Cases
On top of the aforementioned sessions in the main program there are also a few booths where the Scientific-WG is piggy-backing on the generosity of main event sponsor’s through members of the working group:
- Indiana University booth #1000:
- 1030-1100, Tuesday. November 15: Jetstream talk
- 1100-1200, Tuesday, November 15: a series of HPC on OpenStack focused lightning talks
- Presenter: John-Paul Robinson (University of Alabama Birmingham) – Topic: Augmenting traditional HPC services with an Openstack based Cloud
- Presenter: Jeff Adams (Intel Corporation) – Topic: Extending OpenHPC to cloud environments with OpenStack
- Presenter: Blair Bethwaite (Senior HPC Consultant, Monash University) – Topic: Pragmatic approaches to architecting and integrating OpenStack for HPC
- Presenter: Chris Hoge (OpenStack Foundation) – Topic: basic introduction to OpenStack controlling VMs and Bare Metal
- Presenter: Stig Telfer (Cambridge University) – Topic: OpenStack for and high-performance block IO
- Presenter: Lance Wilson (Senior HPC Consultant, Monash University) – Topic: Cryo-electron microscopy image-processing pipeline on OpenStack HPC
- 2:45-3:15, Wednesday, November 16: Jetstream Talk
If you’re coming to SC16 and want to meet up or share notes, please jump on the etherpad. We look forward to seeing you in Salt Lake City! If you are unable to attend the conference, but are still interested in learning more about OpenStack and HPC, visit openstack.org/science.