The NERSC Exascale Science Applications Program, NESAP, is a collaborative effort in which NERSC partners with code teams, vendors, and library and tools developers to prepare for advanced architectures and new systems. NESAP began in late 2014 to help users prepare for the Cori manycore Knights Landing/Xeon Phi architecture. NESAP continues into the 2020's, now targeting Perlmutter, NERSC's first production GPU-based system.
NESAP for Perlmutter
NESAP provides researchers an opportunity to prepare application codes for new architectures and to help advance the mission of the Department of Energy's Office of Science. NESAP partnerships allow projects to collaborate with NERSC and HPC vendors (Cray, NVIDIA, AMD, Intel, etc.) by providing access to early hardware, prototype software tools for performance analysis and optimization, special training, and exclusive hack-a-thon events with vendor and NERSC staff. A subset of projects have the opportunity to work with a postdoctoral researcher to investigate computational science issues associated with advanced architectures. NESAP projects are chosen based on computational and scientific reviews by NERSC and other DOE staff.
NERSC needs to fill postdoctoral and named fellowship positions in support of NESAP for Perlmutter. We will be making a number of NESAP Postdoctoral Fellowship hires and there is a position open for the Distinguished Admiral Grace Hopper Fellowship as well. We encourage applications from NERSC users!
- NESAP for Simulations (N4S): Cutting-edge simulation of complex physical phenomena requires increasing amounts of computational resources due to factors such increasing model sizes, parameter space searches, and inclusion of additional physics. N4S enables simulations to make effective use of modern HPC platforms by focusing on algorithm and data structure development and implementation on new architectures such as GPUs, exposing additional parallelism and improving scalability.
- NESAP for Data (N4D): To answer today’s most complex experimental challenges, scientists are collecting exponentially more data and analyzing it with new computationally intensive algorithms. N4D addresses data-intensive science pipelines that process massive datasets from experimental and observational science (EOS) facilities like synchrotron light sources, telescopes, microscopes, particle accelerators, or genome sequencers. The goal is seamless integration and data flow between EOS facilities and supercomputing resources to enable scalable real-time data analytics.
- NESAP for Learning (N4L): Machine Learning (ML) and Deep Learning (DL) are powerful approaches to solving complicated classification, regression, and pattern recognition problems. N4L focuses on developing and implementing cutting-edge ML/DL solutions to improve scientific discovery potential on experimental or simulation data or improving HPC applications by replacing parts of the software stack or algorithms with ML/DL solutions.
Resources for Projects Selected
- A member of NERSC's Application Readiness team (including members of the Application Performance, User Engagement, Data and Analytics, and Data Science Engagement Groups) to assist with code profiling, optimization, and access to additional resources.
- Additional vendor-provided resources to help with code optimization. NESAP for Perlmutter vendors include Cray, NVIDIA, and AMD.
- MPP hours set aside for code testing, optimization, scaling, and debugging. These hours are shared among the NESAP teams and not added to existing project allocations.
- Access to the Cori GPU cluster (installed in late 2018) for prototyping, debugging, porting, and developing application code.
- Early access to NVIDIA GPU hardware, and early access to compilers, performance tools, libraries, and other software essential for performance profiling and porting.
- Significant MPP hours on the full Perlmutter system (expected delivery 2020).
- Opportunity for a postdoctoral researcher to be placed within your application team (NERSC will fund several postdoctoral researchers and place each with one of the 25 NESAP teams meaning that over half of NESAP applications teams can include a NERSC sponsored postdoc).