Just a little insight - Maybe I'm out in left field playing with daisies, but here goes:
A little on Benchmarks, they are called “SYNTHETIC” for a reason.
Most all reviews use them to show a difference and in some cases it can look like a big difference, BUT if you look at a check using real programs the differences shrink drastically.
... Sequential performance – This is the most often cited matrix – BUT is the least important one for an OS + Program SSD. It is IMPORTANT if the drive is used as a storage drive and you work with LARGE data structures such as the 10 gazzion large photo jpeg/bitmaps, the 1 Gig dot vobs (DVD files) and the 13+ gig Blu-ray files. It is the small 4K performance that is important for an OS + Program drive.
USER Reliability/least user problems to select which of the higher end SATA III drives to Buy – Samsung has a very good rep here.
On “quoted” life expectancy. LOLs. While the tests are somewhat valid they DO NOT reflect real life usage which is going to be a lot less;
... I still think they will outlast there useful life. That is the newest/latest tech will make you want to replace them before they die.
Quick reason why I say the Life cycle test is not valid. Will use a car battery as an example.
When the Battery is getting close to the end of its life. You uses it on a given day and start/stop/restart, NO Problems – it cranks right up. But leave it overnight and it will drain and Walla, dead battery.
The lifecycle test is continued write/delete/write until a cell dies. The NAND cells do NOT get a period of time to allow the charge to decay.
User. Occasionally turns the computer off. Not sure the exact time frame, But if you take a “NEW” SSD and place files on it and take it out of the computer, somewhere between 3 Months and a year the data will be gone. In real life you turn the computer off say overnight. You reach a point that the cell cannot hold the charge. This will occur MUCH sooner than the “life cycle test. Problem is that this type of a test is NOT practical, ie 100 write’s per cell, power off and leave overnight, power on and do another
100 writes per cell ect – This type of test would be valid but take a VERY long time to complete