Could add some substantial speed to conformer generation and MMFF implementation (example implementation [here](https://github.com/NVIDIA-Digital-Bio/nvMolKit/blob/main/examples/conformer_generation_optimization_workflow.ipynb))
