I understand that they're very fast, but I'm not really seeing it. Right now I'm running a Western Digital Caviar Black 1tb and the review in question is of the OCZ Vertex 60gb. Its the very first reviewer, and hes amazed that he can open Photoshop in less than 5 seconds. Seeing this, I tried to open it, timed it (in my head) at around 2 seconds. I've also seen videos of the Crucial RealSSD C300 booting, and its not much faster then my current drive (Although once logged in, it was able to function normally faster).
So whats up exactly? Are SSD's not that fast? Did I have Photoshop cached or something? I had been using it earlier tonight.
EDIT: Tried it with After Effects, opened in around 8 seconds, can someone with an SSD time opening After Effects?
The boot process is not pure, 100%, storage I/O it can only speed up the parts that are.
You're asking for an Apples to Oranges comparison in a completely uncontrolled test.
If you use sleep mode then the contents of the Operating System Disk Cache will remain intact upon resuming your system - it will also use less power than a full reboot in most cases.
If you had been using it earlier that night and have/had ample free* memory thus enabling the OS to get (near) a 100% cache hit ratio instead of having to read it from the physical media then yes you had Photoshop cached.
Task Manager, Performance tab, will show that all (or very close to it) available memory will be used as for the storage subsystem cache. The only exception to this is if not enough disk I/O has occured.
ie: Having 1962MB available memory and only having had 1074MB of cached disk reads/writes since boot up. (Some software explicitly uses uncached I/O as it does not want to affect the cache hit ratio of what is already in the cache, for example CoD4 does this on it's initial load.) [This kind of implies that the Disk Cache has grown to at least 888 MB, the difference between the two figures].
Available memory includes both the disk cache and free memory, as the OS can dynamically shrink to the point of almost not existing. When this happens your cache hit ratio drops to nearly 0% as both reads and write-back caching are negatively impacted.
Free memory is memory that has not been allocated previously, not even by the disk cache. This is often due to not enough disk I/O occuring for the cache to dynamically grow to a size large enough to require all free memory.
There will usually be a 8MB to 64MB discrepency between the totals of Free plus Cached = Available as the Operating System will generally not shrink the disk cache to below 8MB to 64MB in size as doing so would hurt performance more than having to perform swapping of pages in physical memory to and/or from the swapfile(s). (It depends on the cache hit ratio for reads in the last X seconds, the write-back cache benefit, if the storage supports synchronous transfers and so on).
I hope this explanation has been clear and helpful!
I really wouldn't worry about it too much unless you're performing consolidating of Virtual Machines.