1 st Annual SCinet Technology Challenge 2019
1st Annual SCinet Technology Challenge
SC19 Call for Participation
We invite researchers and engineers to submit proposals for the first annual ‘SCinet Technology Challenge’ (TC) intended to highlight closely-coupled operations of sophisticated networking, computing and storage infrastructures in service of advanced science applications. The SC conference series has traditionally been home to cutting-edge developments in high-performance computing (HPC), alongside those in high-performance networking (HPN), storage, and analytics. This challenge, the first of its kind, has the goal of bringing networking, computing and storage together, taking a holistic view to demonstrate the current transformation of science through advanced networked cyber-infrastructure.
The goal of the challenge is to demonstrate that both Networking and high-performance computing resources are essential elements of the cyber infrastructure required to advance modern data-driven science and do so in a creative way that stretches the limits of today’s technologies.
We invite teams of researchers working on the cutting edge of science to provide demonstrations of their advanced scientific applications, while taking advantage of distributed storage and compute resources located on SC floor and elsewhere (university datacenters, national labs, public clouds), linked by ultra-high-performance SCinet network.
Applications must demonstrate some combination of computing and networking and optional data storage, analytics and visualizations, coupled with high-speed data transport that must go across SCinet and partner networks to different locations. Any types of data could be appropriate - streaming or archived. Teams can apply analysis techniques that best suit their use-cases and complement those with vivid data visualizations.
Possible applications and technologies involved include (but aren’t limited to):
- Workflow-driven applications spanning multiple computational resources
- Streaming/IoT applications combining numerous data streams for analysis
- Applications that combine data processing with visualization or AR
- Applications that actively and dynamically orchestrate networked infrastructure to support their needs
Selected teams will provide live demonstrations of their application from the SCinet booth during SuperComputing conference, to be observed by a distinguished jury of academic and industry thought leaders. Winning teams will receive their awards during the SC19 award ceremony.
- June 10, 2019 application submission deadline
- June 30, 2019 TC team notifies selected applicants
- June 30-Aug 31, 2019 SC TC and NRE teams work with individual teams to identify and make available necessary SCinet resources and define the details of each demo
- Nov 19, 2019 TC Day! Teams demonstrate their applications observed by the public and the jury
- Nov 21, 2019 SC Award Ceremony - TC winners presented with awards at SCinet booth
Participating teams must submit a Technology Challenge application (up to 5 pages allowed) through SCinet TC submission web page https://forms.gle/kDwSr9d8QDFo4PQb9 .
The application must include:
- The title of the demo
- The name(s) of the team lead(s)
- The names of all participating organizations
- A short abstract
- A description of the driving application and its scientific significance
- A description of the envisioned demo, including descriptions of data movements and the anticipated timeline of the demo. The description must clearly identify
Computational, storage, visualization and other infrastructure external to SCinet required for the demo.
Desired networking resources that SCinet can provide to connect the demo floor and other required resources to support the demo.
- Approaches and mechanisms for visualizing infrastructure monitoring information to be shown during the demo
Each demo will be alloted 30 minutes, including execution and narrative. Demos are performed from the dedicated TC booth set up near SCinet booth on the show floor.
All demos must include an infrastructure monitoring/visualization component to demonstrate the utilization of different infrastructures, including SCinet
Selection and Evaluation criteria:
Applicant teams will be selected based on a combination of the presented science use-case and the driving application as well as the feasibility of supplying the required infrastructure, both network and (super)compute, to support the challenge demo.
Selected challenge participants will be judged in the following categories:
- Diversity of resource types and geographic distribution, their utilization and degree of orchestration/automation
- Data velocity/variety/volume
- Innovation and the originality of the technical approach
- Human interactivity and effectiveness of the presentation
- Quality/originality of visualization
- The significance of the domain science research problem
An overall winner will be determine by a combination of categories. In addition, one or more awards focusing on specific categories above will also be given out.
SCinet (https://scinet.supercomputing.org/) is a global collaboration of high-performance networking experts who provide the fastest and most powerful volunteer-built network in the world for the SC Conference. For nearly 30 years, SCinet has provided SC attendees and the HPC community with the innovative network platform necessary to connect, transport, and display HPC research at SC from around the world. In addition to high-performance demos, SCinet supports wired and wireless connections for users to maintain connectivity to the rest of the world.
SCinet partners with ESnet, Internet2 and regional networks to provide the necessary high-performance connectivity from the show floor to external resources - campus- and lab-based HPC, HTC, cloud resources, storage resources and scientific instruments.
TC participants’ resource requests will be coordinated through the TC and NRE SCinet teams after the selection is made. Additional information on connectivity options and requests for connections to support TC demonstrations can be made through the traditional SCinet ‘Call for Circuits’ (https://goo.gl/forms/kAu0mKjnfkaw62ap1) .
For questions please contact the SCinet TC team firstname.lastname@example.org .