Test the Scalability of Your PETSc Algorithm on Rescale

The Portable Extensible Toolkit for Scientific Computation (PETSc) is a versatile suite developed by Argonne National Laboratory for solving scientific applications modeled by partial differential equations, widely recognized for its effectiveness in parallel numerical computations. Traditionally, testing new parallel algorithms developed with PETSc on a multi-core computer cluster involved extensive setup and potential administrative challenges; however, using Rescale simplifies this process by allowing scientists to specify hardware and run jobs via an internet connection. The text describes a scalability test conducted on Rescale using a modified KSP algorithm from the PETSc tutorial, which solves a linear system in parallel and logs timestamps for analysis. The test, which used varying numbers of cores, demonstrated that parallel KSP scales efficiently with more cores up to a point, as execution time decreased until 16 cores but increased with 32, attributed to randomness in matrix and vector generation affecting iteration counts. Overall, the analysis showed that the KSP algorithm performs well in parallel execution, and users are encouraged to replicate these tests with different configurations on Rescale.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.