You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Thanks for writing this fantastic library! I had a couple questions I was wondering if someone could clarify:
Is it possible for cuOSQP to operate completely on a GPU alone? It seems that currently the problem is set up in NumPy on the CPU, then passed to the GPU for solving.
Are there any plans to make cuOSQP compatible with PyTorch tensors, CuPy arrays, or other array-like objects that can be stored directly on GPUs? I'm aware that, e.g., PyTorch tensors can be cast back into NumPy arrays, but this requires an extra data transfer step from GPU to CPU when the PyTorch tensor is stored on a GPU.
The text was updated successfully, but these errors were encountered:
Thanks for writing this fantastic library! I had a couple questions I was wondering if someone could clarify:
The text was updated successfully, but these errors were encountered: