SSD Compressed Data

So most people understand Random Data and why 4k / 8k Random results are important to choosing your SSD? If you do not, please go back and read that section for both of our sakes, and the sake of my personal sanity. To keep it in short summary though, 4k & 8k Random Reads & Writes directly represent the most basic & common functions of your Operating System, and thus results will give you the best idea of your "slowest" moments in day to day operations.

Now, at the same time, uncompressed data is very important for a similar regard; as it represents in results, what your SSD will be able to achieve when it's not able to compress data down for more compact storage and better read/write times. For those of you that don't know, the (current) most popular controller out there (Sandforce 2200 series) is infamous (in a good way) for compressing data on the fly when it's written to the drive so that it can thus read the data in the future, at a more rapid rate. Thus, with compressed data (like the small bits of data which make up many components of your OS) you get incredibly quick results.

Now, let's have a small thought experiment for demonstration purposes. When you have a compressed piece of data, the SSD has less to read so it's reading the same amount of information at a quicker pace. On the flipside, if that same data was never compressed to begin with, the SSD will read it at a slower rate. Now, you might ask "what data cannot be compressed" and that is your random data. White Noise is an easy example of data that is completely random and cannot be compressed. As compression relies on finding patterns to map paths to. Thus, large files and certain components do not share enough commonality or have enough patterns to be compressed and thus the SSD will take a longer time to read it.
A simple "experiment" that you can do to show this compression in action is to take a program like 7Zip and compress down 50 copies of the same picture into an archive and then compress down 50 random pictures of the same size into another archive. The 50 copies should be smaller since they should have matching data.