Results

AGI - Analytical Graphics Inc.

09/20/2019 | News release | Archived content

Speed Up Your Trade Studies With Parallel Computing in Analyzer 11.7

There are many engineering problems that you can only solve with iteration. To optimize output values, you're forced to run the same calculation many times with different parameters. Even when you're using engineering software tools, there are several manual steps you must take. Change the input values for each run. Execute the runs sequentially. Record the output for future comparison. It can be time consuming and monotonous.

STK Analyzer provides a set of analysis tools to perform such trade studies, all without scripting. But, like you, Analyzer's automation still executed each step of the analysis sequentially-until now. New in STK 11.7, Analyzer gives you the ability to execute any Analyzer trade study in parallel using STK Parallel Computing.

Parallel Computing can take advantage of the multiple cores on your local computer or a connected STK Scalability cluster. You can configure parallel computing settings in STK on the Parallel Computing page in the Edit Preferences window.

To enable parallel computing in Analyzer:

  1. Click the gear icon in the toolbar to open the Analyzer Options window.
  2. Select the 'Compute in parallel' option.
  3. Click OK.

Analyzer will now use the parallel computing configuration settings that you have defined in STK.

Using Parallel Computing with Analyzer, you can increase the size and complexity of your computations for improved accuracy without waiting longer for them to finish executing. Computing in parallel allows you to take full advantage of existing hardware that may otherwise be idle.

As an example use case, (shown in the video above) we performed a simple parametric trade study to vary a satellite's inclination and compute its coverage over the polar ice cap. In the non-parallel case, it took 105 seconds to compute 10 samples. With parallel computing enabled and using eight local cores, it took 70 seconds to compute 180 samples. That's 33% less time to do 18 times more work. When we used parallel computing configured with an STK Scalability cluster of 60 cores, the same 180 samples took a mere 24 seconds to compute!

Analyzer now joins Coverage, Deck Access, Volumetrics, and Chains in the family of STK computations that can be sped up with parallel computing. The next time you execute a trade study in STK Analyzer, remember to compute it in parallel and see how much faster you get your results!