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
Utilizing signal trees to handle parallelizing multiple tasks from the command buffer. Then we use vulkan's queue submission API to submit these already-prepared tasks to the GPU for rendering. Since Vulkan uses a queue-system way to upload commands to the GPU.
The reason for this is to minimize load-balancing and the amount of wait, locks/unlock when trying to get the data we want to extract.
There are some considerations that I would like to look into which are the following (currently):
- How do we prevent from multiple nodes locking to get the data we want to extract?
- Does selection bias prevent that?
- What are the comparisons of how we traverse up the tree in comparison to the usage of lock-free queues?
The text was updated successfully, but these errors were encountered:
Signal Trees
Utilizing signal trees to handle parallelizing multiple tasks from the command buffer. Then we use vulkan's queue submission API to submit these already-prepared tasks to the GPU for rendering. Since Vulkan uses a queue-system way to upload commands to the GPU.
The reason for this is to minimize load-balancing and the amount of wait, locks/unlock when trying to get the data we want to extract.
There are some considerations that I would like to look into which are the following (currently):
The text was updated successfully, but these errors were encountered: