Any in-depth testing to see if 850 Evo is really not prone to degradation (edited)

[omega3]

Not sure if the 850 Evo has been out long enough, but any signs that 850 Evo is indeed not prone to any performance degradation with stale data?

Yes i know, the 850 Evo has new 3D TLC memory which in theory should be better, but how is it in practice?

If also 3D TLC mem would suffer from any perf degradation, how long would it take to show up cause so far found no complaints but not to many people report about long term usage.

Edit: see my lastest reply below

 

[LTC8K6]

http://www.anandtech.com/show/9158/...-read-performance-fix-coming-later-this-month

http://www.anandtech.com/show/9196/samsung-releases-second-840-evo-fix

 

[LTC8K6]

"Samsung has since moved on to 3D-VNAND in newer drives, which is more stable than very small geometry planar NAND and essentially turns back the clock on NAND scaling issues. So the 840 EVO should be the last Samsung TLC drive to encounter this issue."

 

[corkyg]

Installed a 850EVO as data drive in my laptop last month. No degradation noted.

 

[Prankmaster]

Isn't the 840 better than the 850? it sure is more expensive

 

[LTC8K6]

Isn't the 840 better than the 850? it sure is more expensive

No, the 850 is better.

840 is expensive because it's out of production.

 

[Hulk]

"Samsung has since moved on to 3D-VNAND in newer drives, which is more stable than very small geometry planar NAND and essentially turns back the clock on NAND scaling issues. So the 840 EVO should be the last Samsung TLC drive to encounter this issue."

There you have it. Samsung TOLD US the 850's new technology wouldn't have the problem so that's that. After all they also told to watch out when buying the 840 as it might have a read performance issue. Oh wait, they didn't and still deny it.

 

[LTC8K6]

Well, I have a 250gb and a 500gb 850Evo.

I have no complaints. Had the 250 since January.

 

[omega3]

Given the probs with the 840, hasn't there been any serious test to see if performance degradation with stale files doesn't also happen with the 850 evo.. i don't understand no site has tested this in-depth.

Is it true that the 840evo degradation happened specifically with stale data and that 1 month (?) old data already had the problem or was it longer.. i don't remember

In my situation i actually don't write much to drive.. so most of the files i write would become stale meaning prone to the degradation problem.. therefore i still don't know if buying the 850 evo would be smart in my case? I need more solid advice.

Also, does Magician give you a correct indication of how fast your drive is or do you need 3th party soft?

 

[LTC8K6]

I doubt that 850 owners would keep quiet about it if it were happening.

 

[omega3]

I doubt that 850 owners would keep quiet about it if it were happening.

Yes but where are they

Also, isn't it so that the Magician software has tools in it like wiping the drive in case this would happen.. but then you have to copy everything back ofcourse so not really convenient

 

[LTC8K6]

They are really completely different types of drive.

The 840 being bad doesn't really say anything about the 850.

If I see any slowdowns in the 850 I bought in January, I will certainly be complaining.

 

[omega3]

double post deleted

 

[omega3]

Installed a 850EVO as data drive in my laptop last month. No degradation noted.

Well, I have a 250gb and a 500gb 850Evo.

I have no complaints. Had the 250 since January.

did you test with reading stale files.. as i remember the degradation on the 840 was with reading "old" files and old meaning 1 month or longer since last time accessed

 

[LTC8K6]

To be honest, if I have to test for it to notice it, then I don't care about it.

If I notice it in the normal course of computer use, then I will be complaining.

 

[Crab Balls]

Got four 840s here and haven't noticed any problems yet.

 

[corkyg]

did you test with reading stale files.. as i remember the degradation on the 840 was with reading "old" files and old meaning 1 month or longer since last time accessed

So far, there's no reason to "test." That in itself causes degradation of any SSD. Personally, I view testing and benching as more of a user's ego trip.

 

[dave_the_nerd]

You'd have to take a drive, put some data on it, record a read benchmark, leave it powered on and idle in a system for six months, and then do the read benchmarks again.

If you have time and money, go for it. See you in December.

 

[GlacierFreeze]

You'd have to take a drive, put some data on it, record a read benchmark, leave it powered on and idle in a system for six months, and then do the read benchmarks again.

If you have time and money, go for it. See you in December.

Well the issue does show up in HDTune though...

I think I posted about noticing an issue 6 months before anyone else started posting about it and the reason came out. HDTune graph was horrible. I thought it was a cheap motherboard issue and just ignored it until some figured it out.

 

[LTC8K6]

Well the issue does show up in HDTune though...

I think I posted about noticing an issue 6 months before anyone else started posting about it and the reason came out. HDTune graph was horrible. I thought it was a cheap motherboard issue and just ignored it until some figured it out.

Just to be clear, you are talking about the 840evo, correct?

 

[GlacierFreeze]

Yep.

 

[VirtualLarry]

So far, there's no reason to "test." That in itself causes degradation of any SSD. Personally, I view testing and benching as more of a user's ego trip.

No, it doesn't cause degradation. The SSD speed tester that they are using to test 840 EVO degradation, only reads files, and goes by their file creation / modification date, and benchmarks reading them, and then graphs your files and their speeds in a sort of histogram. I forgot what the exact program is called.

 

[omega3]

You'd have to take a drive, put some data on it, record a read benchmark, leave it powered on and idle in a system for six months, and then do the read benchmarks again.

If you have time and money, go for it. See you in December.

I don't see why you should leave the drive powered.. if you have large files on your system you didn't access in months if you'd copy/paste them from that ssd you would see degradation in speed if the problem still exists.. i remember reading stories like that about the 840.

 

[coercitiv]

No, it doesn't cause degradation. The SSD speed tester that they are using to test 840 EVO degradation, only reads files... snip

Repeated reads can lead to data degradation, which in turn triggers cell refresh.l

The following is quoted from a Intel HET Brief (High Endurance Technology)

With more P/E cycles, read margin shrinks, resulting in both data retention problems and higher read errors beyond the
limit specified by JESD218. This is caused by:

• Loss of charge due to flash cell oxide degradation
• Over-programming caused by erratic program steps
• Programming of adjacent erased cells due to heavy reads or writes in the locality of the NAND flash cell

At the SSD system level, special programming sequences are used to mitigate some of the program state disturb issues that may occur. By using a scheme called background data refresh, the SSD moves data around during periods of inactivity to re-allocate areas that have incurred heavy reads.

 

[omega3]

Installed a 850EVO as data drive in my laptop last month. No degradation noted.

Well, I have a 250gb and a 500gb 850Evo.

I have no complaints. Had the 250 since January.

Does the Magician software have a tool to manually refresh the data by writing it to other cells that also works with the 850 evo.. if that would be the case i would feel more reasurred because in case degradation would happen then you have at least a manual tool to fix that.

Or can you use 3th party soft for that?