By the way, what's the average life of SSDs? Is it similar to HDDs? Are they less reliable?
Overall, while much more expensive, SSDs, due to their lack of physical wear & tear, in the conventional sense of the word anyway (more on that in a minute), are much more reliable than HDDs. You can check out HDDs' expected lifespans here:
http://en.wikipedia.org/wiki/Hard_disk_drive_failure
Flash Memory lifespan depends on the amount of use, and as SSDs are still fairly new, most manufacturers of this type of memory are still gathering date on it's lifespan or MTBF. Common wisdom at the moment seems to be about 10,000 erase/write cycles for the MLC type of chips. (Multi-Level Cell, the most commonly used consumer type of SSD chips - There is also an SLC type of flash memory - Single-Level Cell -, which lasts longer, but is about three times the price of the MLC type; it is mainly used in Enterprise grade SSDs)
The reason SSDs wear out over time, despite there not being any 'conventional' physical wear & tear, is that the cells that make up the chips, have a tendency to hold on to the 'state', 1 or 0, that they are assigned, at any given write cycle; This 'memory' gradually gets stronger after every erase/write cycle, until it becomes so strong that the cell's 'state' can not be changed any longer, at which time your SSD will fail. This is the reason,
reading from the SSD, does not significantly affect it's lifespan, just the erase/write cycles, and this fact plus the blazingly fast access times of SSDs are why the latter are such a no-brainer for holding your OS and assorted software and programs. Check out the boot-time comparisons of a mid-2011 iMac at this link:
http://www.youtube.com/watch?v=cW8Czra3SBA
Wear-leveling techniques, of which 'Trim' is one iteration, can greatly prolong the life of Solid State Memory. A simplified explanation of how wear-leveling works is that Data is written in 'blocks'; by identifying blocks that can be erased and thus re-used, the SSD controller can spread out usage of all available blocks, so as to prevent over usage of some, and very little, or no use of other 'blocks', thereby evening out the wear of all blocks, and preventing premature failure of the SSD, due to heavy over-usage of as little as even a few blocks.