A quick question relating to the write endurance different between the now affordable 'high end' consumer SSD's from the likes of Samsung and the very expensive 'enterprise' class SSD's?
Is this coming closer to a marketing game? Or is there a genuine reason to go for the more expensive offering from the likes of intel?
Other threads on here seem to lack hard evidence or are comparing the cheapest SSD's rather than at least the high end consumer grade such as the Samsung 850 Pro.
Tom
SAS versus SATA is going to be more important in enterprise use cases. Any high-availbility system is going to want dual-ported (SAS) drives, and that's a major differentiator. This is in addition to the presence of a super-capacitor.
The big thing you should look for that makes a drive "Enterprise", regardless of how the drive is labeled and sold, is whether or not is has a supercapacitor to allow the drive to flush the write cache in case of sudden power loss.
Most RAID controllers make certain assumptions that are still based in the old "spinning metal" era of hard drives. These assumptions no longer hold and can really hurt performance in arrays with SSDs. In order to correct for this, SSD drives will often lie to RAID controllers that data in the cache has been flushed before it really happens. In addition, SSDs make heavy use of their cache in order to reduce write wear. This makes the whole array vulnerable to data loss in the case of sudden power loss, where data the RAID controller believed flushed to non-volatile storage was actually still in a cache. SSD drives with the extra supercapacitor protect against this.
There are some consumer-level drives that do have this feature, and some enterprise drives that do not.
Aside from this, enterprise SSDs typically have more raw space on the drive reserved for failed sectors, much like in the spinning metal days, but if you have a good and recent-make (not 1st or 2nd -gen) consumer SSD, you're probably still okay in this area.
