It is just not meant to be a training device so comparing with data Center or developer GPUs is useless. Faster inference for end-users is what is mentioned by

It is just not meant to be a training device so comparing with data Center or developer GPUs is useless. Faster inference for end-users is what is mentioned by Apple and the only use case where this hardware makes sense.