Parallelization & Code Optimization

SDSC has expertise in the areas of software parallelization and optimization. Parallelization refers to the process of taking a serial code that runs on a single CPU and spreading the work across multiple CPUs. This could be done using shared memory techniques to utilize the cores within a single compute node or distributed memory methods that allow applications to run across multiple nodes each with their own local memory. Optimization is the process of making more efficient use of resources and may involve simple steps such as identifying the best compiler options or linking appropriate libraries to significant changes to source code.

For more information about SDSC Parallelization and Code Optimization, please send a request to consult@sdsc.edu.