Understanding RAM and Activity Monitor

I suppose my only follow-up question would be about the compression. Most forums seem to characterize the memory compression as 'free' (making use of an idle core). I wonder how accurate this is?

I think the Compressed memory is what used to be called "Inactive" memory. Inactive memory are pages that have known data or code in them which has been released, but may be asked for again in the near future (for example a program that has just been quit, but may be launched again real soon now; if the program is launched before the data is flushed from the Inactive pool, it will start up much quicker and save both time and power reading it from disk; saving power by not using a disk extends the life of laptop batteries). By compressing the Inactive memory Mac OS X can keep even more idle code and data in RAM than before, and save even more time and power by avoiding the use of the disk. The new Activity Monitor does break out Inactive (compress) from the disk cache, all of which was lumped into Inactive memory in previous versions.

If I were to install 8 or even 16GB or RAM, no doubt the OS would find a way to fill this up. What else would be getting paged in, if there is no overflow at the moment (apart from compression)? Or would the OS just be less judicious about what it writes into the RAM?

It would most likely cache even more file system data. Unused RAM is wasted, and Mac OS X is very good at not wasting your RAM.

The only downside I've seen to the way Mac OS X maintains Inactive RAM is that I think it also avoids flushing modified data back to the file so that the modified data is stored in stable storage. The advantage is that Mac OS X avoid writing to disk until it is forced to and then it writes a lot to concentrate as much I/O into the shorted period of time, then allowing the disk to spin down and save power for as long as possible. The down side is that when Mac OS X need lots of free RAM, it is often forced to do lots of slow writes thus making the Mac appear slow as it is flushing modified data to disk.

The Apple move to using more Solid State Disks in many Mac models, is hiding that down side, but for those of use that are still using traditional disks, the effect is noticeable.

Let's start here:

About OS X Memory Management and Usage

Using Activity Monitor to read System Memory & determine how much RAM is used

OS X Mavericks- About Activity Monitor

Understanding top output in the Terminal

The amount of available RAM for applications is the sum of Free RAM and Inactive RAM. This will change as applications are opened and closed or change from active to inactive status. The Swap figure represents an estimate of the total amount of swap space required for VM if used, but does not necessarily indicate the actual size of the existing swap file. If you are really in need of more RAM that would be indicated by how frequently the system uses VM. If you open the Terminal and run the top command at the prompt you will find information reported on Pageins () and Pageouts (). Pageouts () is the important figure. If the value in the parentheses is 0 (zero) then OS X is not making instantaneous use of VM which means you have adequate physical RAM for the system with the applications you have loaded. If the figure in parentheses is running positive and your hard drive is constantly being used (thrashing) then you need more physical RAM.

Adding RAM only makes it possible to run more programs concurrently. It doesn't speed up the computer nor make games run faster. What it can do is prevent the system from having to use disk-based VM when it runs out of RAM because you are trying to run too many applications concurrently or using applications that are extremely RAM dependent. It will improve the performance of applications that run mostly in RAM or when loading programs.

The numbers provided is simply a running total. There are always pageouts and pageins going on because that is part of memory management. It's the rate or pageouts that is important. When that rate is positive and continuous then the number in parentheses becomes positive and progressively larger. That is the sign of insufficient memory.

Why is this of importance to you?

though I don't understand why virtual memory would be anything other than 0 bytes if Swap Used is also 0 bytes - is this just the amount of disk space set aside by the OS should it be required

Because virtual memory has nothing to do with disk. It is how the hardware manages the address space assigned to processes. A process has a virtual address space from 0 to n. That address space may have holes in it which will never have RAM associated with it, but it is still part of the process' address space. The address space may include a large memory mapped file, so that address space is large, but the data for that is backed by a real file, not swapfiles. Shared library/framework code is part of many address spaces, but it is still code residing in a regular file on disk, and again not swapfile space needed for that. Also that virtual address space is multiplied by the total number of programs sharing the library/framework creating larger virtual address spaces than you expect.

The size of your virtual address space if mostly something no one cares about except a very few operating system developers, and they tend to be the ones that are actually maintaining with the code that deals with virtual memory.

Pageins: 285651. Pageouts: 35809.

How can I have pageouts above 0 if Swap Used is 0 bytes? Moreover, Pageouts/Pageins > 0.1, which lots of people quote as a good rule of thumb for judging whether your RAM is overstretched. So what's going on? How can the system be writing pageouts without using any "swap". My "Memory Pressue" meter is nice and green and low but I'm not sure I trust this as it's unclear how much of an abstraction it is.

Because at that moment in time, no processes had any of their data written to the swapfiles. Sometime in the past a process (or processes) needed to be paged out. The vm_stat value is cumulative from the last boot.

Also a memory mapped file that is modified in memory will be written back to the file using pageouts, but that would not affect swapfiles.

I'm thinking about buying RAM and just generally want to understand my system better and make sure I'm getting the most out of it.

Start an Applications -> Utilities -> Terminal session, and run the command:

sar -g 60 100

which will report the pageout activity once a minute for 100 minutes (adjust the numbers to suit your tastes).

Now use your Mac under the load conditions where you think you might want more RAM. Come back later and look at the sar output. If you mostly have zeros, with maybe a small blip now and again, you are good. If you some a few small pageout blips, unless you notice the system performance is bad during those blips, then your are still good. If you have very large pageout numbers, or sustained pageouts, then you may benefit from additional RAM, especially if you can associate poor performance with paging activity.