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
Is your feature request related to a problem? Please describe.
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]
No Describe the solution you'd like
A clear and concise description of what you want to happen.
When running diffusion simulation, multiple running is required to obtain a mean value on node influence, this process can be written for CUDA, and it will speed up when using GPU. Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.
No considered Additional context
Add any other context or screenshots about the feature request here.
Please add support for GPU CUDA.
The text was updated successfully, but these errors were encountered:
Is your feature request related to a problem? Please describe.
A clear and concise description of what the problem is. Ex. I'm always frustrated when [...]
No
Describe the solution you'd like
A clear and concise description of what you want to happen.
When running diffusion simulation, multiple running is required to obtain a mean value on node influence, this process can be written for CUDA, and it will speed up when using GPU.
Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.
No considered
Additional context
Add any other context or screenshots about the feature request here.
Please add support for GPU CUDA.
The text was updated successfully, but these errors were encountered: