Kombyne™ works with your simulation solvers to extract only “key” simulation results to significantly reduce run times and data storage. These reductions enable extraction of results data more frequently to ensure that all appropriate data is available. Kombyne™ works as a co-process with your simulation solvers and can adjust simulation parameters as the simulation is running. Kombyne™ is based on extensive experience of large scale simulations at Intelligent Light to enable you to:
- Enhance throughput and reliability of HPC simulation by efficiently managing complex HPC workflows
- Increase simulation workflow confidence with consistent smart key simulation data extraction
- Increase productivity via standardized open extracts instead of proprietary files
- Reduce the risk of simulation-based design errors through more frequent extraction of key simulation data
- Reduce costs by migrating post-processing to automated batch workflows
- Reduce HPC resource requirements for simulation
Enterprise Open-source Tools to Help You Use In Situ processing, In Transit processing and Solver Steering
In Situ approach extracts key data from the solver executing on the HPC removing the need for large result files
In Transit approach moves simulation data processing to different computing resources for better simulation performance
Solver steering allows adjustment of solver parameters while the solver is running based on extracted data
“In this approach the Rolls-Royce flow solver Hydra was instrumented with the Kombyne software which allows flow solver parameters, such as the prescribed exit mass flow, to be changed while the solver is running and the system output parameters to be monitored and processed automatically. This technique avoids the need to submit separate jobs for each operating condition and treats the CFD simulation like a virtual experiment, where settings are changed by the analyst while the “virtual engine” is running.”
AIAA 2022-1723 SciTech 2022
Affordable
Provided via an annual subscription without user count restrictions and in source code form.
Efficient
Built to be memory efficient and easy to build. Free from dependencies like vtk and adios2.
Scalable
Unique third-generation in transit architecture scales easily with your solver runs.
Domain Experts
Built by a team of experts in CFD solvers, visualization and in situ technologies.
We Can Help You Get to Smarter Simulation Data Faster
- Increases simulation workflow confidence with consistent smart key simulation data extraction
- Integrates with IntelliTwin™ to obtain extracts and metadata
- Based on decades of complex real-world simulation experience
Additional Resources
Kombyne™
Industrial and Biomedical CFD Workflows Enhanced via Coprocessing for Knowledge Capture and Computational Steering,
Steve M. Legensky, Earl P.N. Duque, Dave A. Amels and Brad Whitlock, Marcus Meyer, Paolo Adami, Axel Gerstenberger and Rene Thümmler, Eleventh International Conference on Computational Fluid Dynamics (ICCFD11),
Maui, Hawaii, USA, July 11-15, 2022