SDSC to Help Deliver UHPC Extreme Machine in 2018

Print Friendly, PDF & Email

Exascale computing discussions often center around when and how, but now it looks like we might just know where. This week the San Diego Supercomputer Center (SDSC) announced that it will provide expertise to a multi-year technology investment program to develop the next generation of extreme scale supercomputers.

As part of the DARPA’s Ubiquitous High Performance Computing (UHPC) program, SDSC researchers are working on applications for the project , which being lead by the Intel Corporation. The multi-phase project calls for a full system design and simulation to be completed in 2014. Phases 3 and 4 of the project, which have not yet been awarded, are expected to result in a full prototype system sometime in 2018.

We are working to build an integrated hardware/software stack that can manage data movement with extreme efficiency,” said Allan Snavely, associate director of SDSC and head of the supercomputer center’s PMaC lab. “The Intel team includes leading experts in low-power device design, optimizing compilers, expressive program languages, and high-performance applications, which is PMaC’s special expertise.”

SDSC will be Booth 2239 at SC10. Not to be missed is their session entitled: From Data to Discovery: Lessons on Using Data to Connect Events in Our Environment.

ABSTRACT:
A growing challenge teachers face is the need to deliver curriculum anchored in real-world scientific research. This session introduces teachers to some of the growing number of sources for data-oriented curriculum and analysis tools, all freely-available and teacher- and student-friendly. We begin with a hands-on introduction to the SDSC Discover Data Portal, a collection of freely-available lessons. Participants work with scientific data from fields including astronomy, earth science, marine science, and environmental science, stepping through introductory lessons using simple data analysis tools such as Excel to more advanced lessons using analysis tools such as Octave (a free Matlab-like program). Teachers will learn how to generate their own Google Map and Google Earth “Mashup” visualizations of earthquake data, climate models, and gulf oil spill simulations. Teachers will then explore some of the educational resources within the TeraGrid Science Gateways, including the LEAD Gateway on climate modeling and NanoHub, the nanotechnology Gateway.