The Virtual School of Computational Science and Engineering (VSCSE) helps graduate students, post-docs and young professionals from all disciplines and institutions across the country gain the skills they need to use advanced computational resources to advance their research. The VSCSE deploys conventional collaboration technologies in unconventional ways to create a national-scale virtual classroom that provides multiple high-quality audio and video channels for speakers, remote audiences, and various forms of content of immediate educational value to students.
“The largest video conferencing class in the country and one of the most productive weeks of your graduate career”, Wen-Mei W. Hwu, Professor of Electrical and Computer Engineering and Principal Investigator of the CUDA Center of Excellence, University of Illinois at Urbana-Champaign.
Seats still available for the Summer School workshops:
- Science Cloud Summer School, July 30 – August 3rd
- Proven Algorithmic Techniques for Many-core Processors, August 13th – 17th
The Science Cloud Summer School targets education and training of graduate students and the fostering of a community around a topic that has increasing interest and relevance: the use of cloud computing technologies in science – including infrastructure-as-a-service and platform-as-a-service. Because cloud computing systems and technologies provide a considerable departure from traditional models and evolve at a rapid pace, this course will provide a basis for students to immerse in a focused, intensive curriculum to learn fundamentals and experiment with these technologies in practice. We will cover topics of interest to students with both application and computer science focus.
The Proven Algorithmic Techniques for Many-core Processors: Studying many current GPU computing applications, we have learned that the limits of an application’s scalability are often related to some combination of memory bandwidth saturation, memory contention, imbalanced data distribution, or data structure/algorithm interactions. Successful GPU application developers often adjust their data structures and problem formulation specifically for massive threading and executed their threads leveraging shared on-chip memory resources for bigger impact. We looked for patterns among those transformations, and here present the seven most common and crucial algorithm and data optimization
techniques we discovered. Each can improve performance of applicable kernels by 2-10X in current processors while improving future scalability.
Registration fees for each course are $100 and help offset the logistics and hospitality costs that the host sites incur in order to make this exciting program happen. Registration is open at the VSCSE Hub. More information on the course topics and instructors can be found on the VSCSE Website.
Students are required to provide their own laptops for accessing HPC resources for hands-on laboratory sessions.