#M1Pro and #M1Max certainly sound like they have UMA GPUs with similar compute performance and on-chip bandwidth to high end discrete GPUs right? Let’s see