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
Currently, we have mechanism to to detect cluster description with tt_ClusterDescriptor::get_cluster_descriptor_file_path() for devices that are attached to host. We also have tt_ClusterDescriptor::create_mock_cluster function that creates mock version for specified device ids and architecture.
Minimum what we should have is function that will create cluster descriptor for simulation device based on simulator directory.
Since we will have function for "real" devices connected to host and also function for simulaton device, not sure if it makes sence to have static methods in tt::umd::Cluster and tt_SimulationDevice that would be logically associated with those functions...
The text was updated successfully, but these errors were encountered:
Currently, we have mechanism to to detect cluster description with
tt_ClusterDescriptor::get_cluster_descriptor_file_path()
for devices that are attached to host. We also havett_ClusterDescriptor::create_mock_cluster
function that creates mock version for specified device ids and architecture.Minimum what we should have is function that will create cluster descriptor for simulation device based on simulator directory.
Since we will have function for "real" devices connected to host and also function for simulaton device, not sure if it makes sence to have static methods in
tt::umd::Cluster
andtt_SimulationDevice
that would be logically associated with those functions...The text was updated successfully, but these errors were encountered: