To me, that is still unclear. If you read between the lines of the article link, it may refer to cache memory. IOW, stuff can be offloaded to the VM file on a HDD or SDD, or it can be offloaded to L3 or other cache (not sure if L3 is still used). Since cached data is acquired as fast as, or even faster than RAM (due to how fast the bus may or may not be) there is an advantage to offloading data to a cache. Maybe this is what we are talking about.What exactly is that advantage?...
Ironically, the G3 chip originally processed significantly faster in the real world than the G4 chip because the G3 development team concentrated on having a lot of L3 cache while the G4 team did not. This was difficult to explain, and resulted in a lot of chips being plowed under to cover this fact up (source: Leo LaPorte on MacBreak Weekly).
So if we assume that this is what they are referring to as "Inactive" memory. and I am purely speculating here, that would at least explain why it may have an "advantage" that free memory (memory unused and sitting in RAM) might not.
The obvious question then becomes "what does 'purge' do?". Does this mean that when you invoke a purge command that it blows that memory out of cached status and back into free memory?
If so, then it makes sense not to invoke it if you expect to reload recent documents or apps. And it seems then not to matter if you don't expect to. And even then, any "advantage" seems pretty minor.
Especially in the era of SSDs. It would seem that accessing RAM, while significantly faster than accessing a HDD, might not be all that different than accessing an SSD. That means the "advantage" of cache memory is even less than it would be with a standard HDD, and the same applies to how much virtual memory there is being used, and to the page ins and page outs.
Can anyone confirm or deny any of this?