[quote author=“gr00vy0ne”]Of course, playing devil’s advocate…why would you want to? It’s lot like we’re being limited in any way by the current implementation. The TR can access the maximum amount of memory the chipset provides and pre-allocating the memory probably won’t make things any faster anyways…otherwise Intel probably would have provided a way to tune it manually.
2 reasons that I could think of…
1. Just because.
2. I am not 100% sure if it works the way it should…for example, within a game, does it release the memory ontime? IMO it prob. waits until the buffer is full, and then opens more memory, leaving the game skipping for a minute. Unless it adds 2MB on top of whatever it’s using at all times.
Things to think about.
Erik