"HPC (High Performance Computing) users face a range of hurdles to share their data", stated SDSC Visualization Group Leader Amit Chourasia, also SeedMeLab's Principal Investigator (PI). "Their collaborators may not have adequate context of the computed data, or may not be able to find, access, or fetch the data from HPC system. Collaborators also often have a pressing need to repeatedly find, access, and review data products in a fast-paced environment. These hurdles inevitably create a bottleneck at best, or in the worst case can cripple the scientific discovery process and also burden the principal HPC user to devote considerable time and effort to develop ad-hoc and poorly maintained data sharing methods rather than focusing on the research."
HPC Share solves these hurdles by letting users accelerate the research pace and information exchange via a ready-to-use infrastructure. Its key capabilities include easy data transfer, accessibility and sharing via a web browser on any device, an ability to add annotation to any file or folder, discuss and visualize tabular data. HPC Share is available to all users of SDSCs Comet supercomputer, with a potential expansion to SDSC's Triton Shared Computing Cluster (TSCC) and its Expanse supercomputer slated to enter production later this year.
"HPC Share bridges a major gap in our current cyberinfrastructure by offering a turnkey system that eliminates barriers among collaborating researchers so that they can quickly access and review scientific results with context", stated SDSC Director Michael Norman. "Users will benefit from reduced complexity, ubiquitous accessibility, and more importantly rapid knowledge exchange."
HPC Share is powered by SDSC's open-source SeedMeLab software that was developed with support from National Science Foundation (NSF). Its built-in web services, coupled with an API extension, make it a versatile platform to create branded data repositories for small research groups to large communities or integrate with existing research data flow while also serving as an important stepping stone for researchers to realize FAIR data management in practice.
Researchers interested in learning more about HPC Share and SeedMeLab are invited to participate in a free webinar on March 19 that will cover the use of the data management content system on SDSC's Comet supercomputer.
"We have set up a dedicated instance of SeedMeLab for our HPC users", stated Amit Chourasia, who will lead the webinar. "We will provide hands-on training to move, share, and present your data to and from Comet that will organize and store your data in one place while making it accessible to other researchers." Details of the webinar can be found online .
In addition to Amit Chourasia, the SeedMeLab project includes SDSC Director Michael Norman as co-PI and David Nadeau as a technical architect with SDSC. Master's and undergraduate interns at UC San Diego's Computer Science and Engineering Department, as well as regional high school students, have assisted in the project with prototype extensions and comprehensive quality assurance of the software. SeedMeLab is supported by NSF grant #OAC-1443083.
More information is available at the SeedMeLab website.