GitLab

* Improve selection of architecture to compile for

* Fixed compilation errors on older CUDA versions

* Changes name of NVRTC program

* Adds launch bounds for findInteractingBlocks

* Replaces launch bound parameter with named constant
Co-authored-by: David Clark <daclark@nvidia.com>

Implemented setting random seed for reference platform, as well as adding it as an option to BrownianIntegrator and AndersenThermostat.

Implemented dynamic loading of platforms from DLLs.  Also fixed a lot of compilation errors and warnings under Windows.

Modified the initialization order of ForceImpls to prevent them from trying to initialize themselves before the Platform had been chosen