If in fact at 16MB for every block, this would signify 64MB of on-chip generic cache for The entire SoC to take advantage of. Outside of the apparent GPU works by using, I do question exactly what the CPUs have the ability to accomplish with these kinds of gigantic memory bandwidth assets.was discovered which the AJM take a look at benefits didn't