Parallelizing STK

Parallel computing can support increasing complexity and scale of analysis. Systems Tool Kit (STK) offers different methods for parallelizing tasks.

1. STK’s Native Parallel Computing Capabilities

STK comes with built-in Parallel Computing capabilities, designed to distribute computationally intensive tasks across multiple cores on a local machine or a cluster. This capability simplifies the parallelization process for a specific set of analysis tasks, making it accessible without needing additional code development. The analysis in STK that can be parallelized include:

• Coverage and Figure of Merit (FOMs): You can compute high-resolution coverage grids fast.

• Volumetric Analysis: Run 3D Coverage across multiple cores.
• Chain Access: Processes multiple chain access definitions in parallel.
• Deck Access: Process Access from a single object to a whole satellite database, star database, city database or the Satellite Collection object.
• Recording Movies and STK Analyzer: Parallel rendering and analysis speed up operations like creating movies and running complex simulations.
• EOIR: Parallel creation of Synthetic Scenes.

This out-of-the-box parallelization is convenient but limited to specific tasks and a fixed method of parallelization.

2. STK Scalability Extension: A Flexible Approach to Parallelization

The Scalability Extension provides a more flexible approach to scaling analyses across multiple cores, clusters, or even cloud environments. It allows for custom parallelization of tasks beyond what the native Parallel Computing capability supports.

Key Features of Scalability Extension:

• Custom Workflows: Unlike the native parallel capabilities that handle predefined tasks, Scalability Extension enables you to parallelize any task by automating workflows via the STK API or external tools.
• Cross-Core and Multi-Server Execution: You can distribute computational tasks across multiple cores or servers, ideal for large-scale simulations and analyses.
• Integration with AGI Products: It integrates with other AGI tools like ODTK, allowing for additional distributed workflows.
• Development Required: Developers can use Python, Java, or .NET SDKs to automate and manage the parallelization of custom analyses.

• STK API
  1. Object Model: https://lsas-tec.freshdesk.com/support/solutions/articles/150000180790-getting-started-with-object-model-for-stk
  2. Connect: https://lsas-tec.freshdesk.com/support/solutions/articles/150000174167-getting-started-with-connect-commands-for-stk
  3. Python Specific: https://lsas-tec.freshdesk.com/support/solutions/articles/150000185087-python-with-stk-01-initial-connection

3. Licensing Considerations for Scaling STK

When parallelizing STK, consider the licenses required.

• Native Parallel Computing: STK allows for parallel computing locally without any additional licenses.
  1. For STK Pro License, you can use up to 8 cores.
  2. For STK Premium (Air or Space), you can use up to 16 cores.
  3. If you want to use additional cores. The HPC license can be used to add additional cores. Parallel Computing can use an Ansys HPC license, and the number of cores available depends on your specific license tier.
• Scalability Extension: Due to the flexibility and the development work required, there is no specific license for this scalability as STK does not know it has been scaled. A full STK Engine license is required but LSAS provides some options. As each core running the STK analysis, you will need a STK Engine License. The intention is NOT to require users to purchase a full-priced STK Engine License for each core. Please work with your LSAS account manager to provide pricing information for the necessary STK Engine licenses.

Conclusion

By leveraging STK’s Native Parallel Computing and the Scalability Extension, you can significantly reduce analysis times and increase the fidelity of your simulations. These parallelization techniques offer the flexibility to scale operations on a local machine or across a distributed cluster. The ability to automate workflows through APIs further enhances the potential to parallelize complex, custom analyses, making these tools versatile in modern space applications.

For further technical support or to explore how to implement these solutions, contact our support team at support@lsas-tec.com.