Special Session on:
Applying High-Performance Computing for Scientific Discovery within Real-World Problems

Conveners:

Glenn Hammond, Pacific Northwest National Laboratory
Haibing Shao, Helmholtz Center for Environmental Research – UFZ

Description:

Advances in high performance computing have enabled the development of massively-parallel simulation frameworks that employ thousands of processors simultaneously. Although the development of these frameworks imposes challenges such as data structures and numerical algorithms that need to be adapted to maximize scalability and fully utilize the greater computing capacity, their deployment enables the potential for accelerating scientific discovery, such as 1) coupling increasing numbers of physical and biogeochemical processes with high-fidelity discretizations; and 2) executing large ensembles of realizations to address uncertainty of stochastic parameters within natural and engineered systems. This session seeks presentations that demonstrate the utility of high performance computing to engender new scientific discovery on real-world problems. These include, but are not limited to:

• System-scale modeling of field and laboratory experiments

• Performance assessment for engineered systems

• Sensitivity analysis and uncertainty quantification with stochastic methods

Demonstration of advanced numerical methods is encouraged, but should be presented within the context of the new scientific discovery enabled by the method.