Personally, I buy SSDs because I want to improve performance. In my case, I find it pointless to write as little as possible to the SSD just to extend its life from 30 to 100 years if performance has to suffer.377Ohms wrote: ⤴Sun Aug 12, 2018 8:44 am I use simulation programs that use the likes of Finite Element Analysis (FEA) and SPICE based circuit simulations. These applications can have huge dynamic memory and storage appetites. As such, disk caching goes on all the time. I've seen warnings about these applications rapidly wearing out SSDs. (I know from practice they're really hard on HDDs.) Best practice seems to restrict them to caching on rotating storage only. It sounds plausible, but then again there seems be a lot of different opinions out there when it comes to the subject of SSD lifespan and performance.
There have been plenty of endurance testing done by third parties on SSDs that I'm relatively certain the # of P/E cycles at my given workload is more than sufficient. Indeed, I've replaced SSDs due to capacity and falling prices before any of them have even used 100 P/E cycles.
http://www.xtremesystems.org/forums/sho ... nm-Vs-34nm
https://us.hardware.info/reviews/4178/1 ... -20-6-2013
[attachment=42861]