View Full Version : ML Ram Usage
Jul 17, 2012, 03:51 AM
Can anybody talk about ML ram usage?
Better than Lion? More than Lion?
I heard people crying telling that ML would need at least 16Gb (which is obviously retarded) and others tell that it will use much less memory...
Examples like ram used after boot would be welcome.
Jul 17, 2012, 04:21 AM
According to activity monitor, 800MB active, 300 wired. That's with Safari and Activity Monitor open.
Jul 17, 2012, 05:21 AM
Activity Monitor says:
So about 1.8GB used. Same as Lion for me.
Jul 17, 2012, 05:58 AM
I only have 4GB since I have an old MBP so I really need my free memory.
In Lion, I would get LOTS of page outs. I would really notice that once I minimize an application like Mail, and then use Firefox or something for some time and then open up Mail again, it would take about 5-8 seconds to just load everything back up.
In ML though, the first thing I have to say is that I always have more free memory available which is fantastic. Everything feels more fluid too. I can definitely tell they did some serious work on memory management. I'm even using Safari instead of Firefox because it uses small amount of memory compared to Firefox for now. I have 22 tabs open right now and it's only using about 900MB total which is not too bad! Firefox takes up about 1.2GB for the same websites.
I'm loving ML so far. To me, it feels just like the update from Leopard to Snow Leopard. Disappointing amount of new features but lots of welcome improvements under the hood.
For some stats: for 6 days and 15 hours of uptime, I have 60.56 GB page ins and 2.66 GB page outs. Now that's impressive.
Jul 17, 2012, 08:48 AM
Page-ins always happen even if you have tons of RAM available. OS uses memory mapping to run executable binaries. The first time a page is accessed, the content will be loaded from the mapping file, that's one page-in.
If you see a lot of page-outs, that means you have run out of RAM pretty badly. Either the OS is doing a bad memory management job (like Lion, Windows Vista), or you have little ram to support your usage patterns.