Tools and Software
To explore unique phenomena associated with clouds of collapsing bubbles, LLNL scientists and collaborators performed the largest-ever fluid dynamics simulation on the Sequoia supercomputer.
Livermore Computing (LC) develops and implements systems software that helps the science get done. An international team of scientists was in the midst of simulating a collapsing cloud of 15,000 bubbles using Lawrence Livermore’s massive Sequoia supercomputer when the calculations suddenly stopped. The team turned to a Livermore software tool called the Stack Trace Analysis Tool (STAT) to locate which of more than 6 million computing threads (calculations) was causing the problem. Within a few minutes, STAT traced the hangup to a particular microprocessor core (computing engine), and the researchers were able to fix the problem. The team went on to complete the pioneering simulation and win the 2013 Gordon Bell Prize for outstanding achievement in high-performance computing.
Developing and supporting tools that help Livermore supercomputer users—scientists and engineers—run their simulations and calculations efficiently and effectively is the task of organizations such as LC’s Development Environment Group (DEG), the group that created STAT. LC computer scientists collaborate with other LLNL organizations, universities, and computer hardware and software companies to build and refine reliability, performance analysis, and visualization tools; compilers; and many other software components. While these tools rarely share the spotlight with the massively parallel science codes running on Livermore’s supercomputers, they are helping to enable great science.
For more on LC tools and software development:
- DEG team page
- Software and Development Environment Overview
- Supercomputing Tools Speed Simulations, Science & Technology Review magazine, 2014
- Lightweight, Scalable Tool Identifies Supercomputers’ Code Errors, Science & Technology Review magazine, 2011