Is the underlying NAND not the limiting factor here? Or are they just being lazy and not giving real data? It doesn't seem to have any relation whatsoever to the 3,000 program/erase cycles they're claiming on their 20nm NAND.
I have an application that could benefit from the random performance of an SSD and I was hoping with normal program/erase cycle based lifespan expectancy that the 960 GB would last years under my workload (about 6-8 MB/s writes continuously) but if 72TB is the real life expectancy I'll burn them out far too quickly to make the cost feasible.
Viper GTS
I have an application that could benefit from the random performance of an SSD and I was hoping with normal program/erase cycle based lifespan expectancy that the 960 GB would last years under my workload (about 6-8 MB/s writes continuously) but if 72TB is the real life expectancy I'll burn them out far too quickly to make the cost feasible.
Viper GTS