Get the proper physx multi thread fix and any modern cpu made in the last 15 years basically will blow passed any nvidia gpu accelerated physx... there's a reason nvidia clamped down on physx in 2009 onward more and more when modders figured out how to lift the restrictions nvidia placed on it. It was limited to a single thread and of course performed poorly, but with the modification, allowed for any additional free cpu resources to be used, heavily, but kept it from pegging the cpu at 100% to avoid issues, the result of which is anyone with a Intel i7 900 series or even a modern i5 when they arrived, could accelerated physx as fast or faster than nvidia's own gpus at the time used as a secondary even. People with radeon gpus were playing many games at higher sustained frame rates than nvidia's offerings, specially through those years when nvidia was repeatedly producing garbage gpus. radeon gpus were running faster, but it has to use the modded physx file since nvidia got butthurt and decided to hardcode single thread support for physx, meaning you have to drop the older modded physx in place. Once that's done, any quad core or better, specially modern cpu will WRECK any nvidia gpu doing terracore physx on chip. This has always been the case since it was done, IT'S PRECISELY why nvidia was angry about the fact.