[Gmat-developers] Greetings and some questions

[Harris, Ben]

Hello everyone,

 

My name is Ben Harris and I work at Applied Physics Laboratory. I am the navigation lead for the Interstellar Mapping and Acceleration Probe. We plan to use ODTK for our OD but I am investigating alternatives to double check our results from the ODTK. Also I am looking for ways to accelerate our mission Monte Carlos. I have experience with ephemeris generation, C++, CUDA and open source, in particular the GPS Toolkit (GPSTk, www.gpstk.com).

 

I am super impressed by GMAT’s capabilities and where those are going. I anticipate I will be utilizing GMAT and will eventually contribute back at least with bug fixes.

 

Given that the base code for GMAT is in C++, it strikes me that it could be a great place to implement Monte Carlos that take advantage of GPU acceleration. I have a few questions regarding using GMAT in this regard:

 

1.      Are there any plans to incorporate a MC framework within GMAT? So users could parallelize GMAT processing without external process control.

2.      Has anyone ported GMAT’s C++ code to CUDA?

3.      In a similar vein, has anyone built GMAT with a LLVM (low level virtual machine) compiler?

4.      Is there a notional schedule for the release of r2021?

 

Thanks for whatever guidance you can share.

 

Regards,

Ben