Support for adding additional GPUs to task pools, so that they can be used to offload secondary compute or render tasks. For example, pathfinding/AI or texture generation tasks.
A mechanism may also need to be implemented to simplify duplicating, copying or transferring resources between GPUs, since resource cannot be directly shared between different GPUs.
A mechanism may also need to be provided for specifying which types of tasks each additional GPU is allowed to process, as a developer (or user) may want to assign lower priority tasks to less powerful GPUs.
Support for adding additional GPUs to task pools, so that they can be used to offload secondary compute or render tasks. For example, pathfinding/AI or texture generation tasks.
A mechanism may also need to be implemented to simplify duplicating, copying or transferring resources between GPUs, since resource cannot be directly shared between different GPUs.
A mechanism may also need to be provided for specifying which types of tasks each additional GPU is allowed to process, as a developer (or user) may want to assign lower priority tasks to less powerful GPUs.
This will rely on: