I am trying to run distributed training with TF-Metal and two M1-Ultra Studios. On connecting them via thunderbolt cable, I am able to see the device in the "About Mac" column but TensorFlow doesn't pick that GPU up.
Commands I am using:-
mirrored_strategy = tf.distribute.MirroredStrategy()
Output:-
INFO:tensorflow:Using MirroredStrategy with devices ('/job:localhost/replica:0/task:0/device:GPU:0',)
TF-Metal version: 2.9.2