Any Techreport-esque TLC torture tests?

[PliotronX]

I know the drives they had tested were some upper crust consumer based models that were MLC but I was wondering if even through happenstance somebody has seen hockey stick sector reallocation with TLC based drives. Have they exceeded expectations as well in p/e?

 

[Glaring_Mistake]

Hockey stick sector reallocation?

Anyway, they are running similar tests on TLC drives at Nordichardware.se.

The first to die was (expectedly) the SanDisk Ultra II (500 P/E) that managed to get up to 524TB in writes but that was at a WA of 0.5 due to the SLC-cache and compression it used meaning it reached 1000 P/E rather than 2000.

The second was the 850 EVO (2000 P/E) after 1000TB and about 4000 P/E.
After reaching 1000TB they left it unpowered a while with a folder of files and when it was time to test the integrity of those files with a hash-check it slowed down before dying.
However since they did not test it with hash-checks before that we do not know exactly when it was really starting to develop issues.

The third (which has yet to die) is the 840 EVO (1000 P/E) now at 1800TB and 6600 P/E.
Unlike the 850 EVO this one did survive being unpowered after 1000TB written.
Even though it has yet to die it has tens of thousands of Uncorrectable Errors and it's write speed has slowed to a crawl.

The fourth (which is fairly new) is the ADATA SP550 (1000 P/E) which has only been through 90TB or 350 P/E.

Think 840 EVO is the one to develop most reallocated sectors with an entire nineteen of them.

 

[Tjalve]

http://www.nordichardware.se/test/samsung-840-evo-nar-1-600-tb-i-nhws-ssd-dodartest.html
The site is written in Swedish but the graphs are in English and shows the progress of the drives.
840 Evo is just a matter of time befor it stops working. The SP550 is still kicking and should be good for at least another two months.