Two Samsung 840 Pro's & 1 840 have failed, 4 at [H]ardOCP (Early review sample SSD's)

[katkah]

I purchased mine a couple of days ago and received today. I think it EXTREMELY ironic that the 2 failed exactly coinciding with their release to the public. I purchased the US version and both posters seem to be posting from elsewhere. I didn't open waiting for some real life reviews but based on the previous version and the additional year of research I can't wait to see what consumer level users have to say. It is reported to come with a game (assasin III), but I'd rather have lower prices. I guess that they are trying to hook gamers on the SSD experience.

[Puffnstuff]

I cloned my boot drive the my 840 pro this afternoon but haven't installed it yet. I didn't receive any games with my drive.

[katkah]

Originally Posted by Puffnstuff

I cloned my boot drive the my 840 pro this afternoon but haven't installed it yet. I didn't receive any games with my drive.

It's being reported everywhere:

http://www.engadget.com/2012/11/07/s...h-ssd-840-pro/

Look through your paperwork as it is just a code and a link to download the product.

I'm going from a 7200 to the ssd so I am doing a clean install to ensure proper settings, but will definitely wait before I open it. Not actually scared but I want to hear others experiences with everything.

[Puffnstuff]

Getting back on topic this is the only place that I see reporting a failure of this drive. I'll swap mine in soon. I also double checked the the contents of my drive and no game code or reference to a free game.

[videoclone]

Originally Posted by Zap

Do you live near a Micro Center? I hear the Samsung 830 256GB is $149.99 in store, and the 128GB is $69.99, no rebate pricing.





I think graphics cards are fine as well as CPUs, but yeah SSDs and motherboards often seem iffy before the first round of firmware updates.

Im in Metro, Sydney, Australia.. Cheapest i've seen it for is $195AU ... its not as good as $149.99US but its not bad either. Sandisk's can be had for $170

The exchange rate is $1 AU buys $1 US so pretty much the same apart from tax and larger market making the cost cheaper for the US

[videoclone]

Originally Posted by jwilliams4200

Did you test either of the failed units with a meter to see if the power delivery circuit was providing power to the ICs?

Did the second unit fail with the exact same symptoms as the first unit? What exact I/O was the second unit subjected to when it failed?

He sent the drives back to samsung

[jwilliams4200]

Originally Posted by videoclone

He sent the drives back to samsung

Obviously. Did you have a point?

[Coup27]

Originally Posted by katkah

I purchased mine a couple of days ago and received today. I think it EXTREMELY ironic that the 2 failed exactly coinciding with their release to the public. I purchased the US version and both posters seem to be posting from elsewhere.

Would you care to expand these comments? They suggest that Anandtech has purposely given Samsung some bad publicity at the same time as a product release.

The second statement is just thick. Do you think Samsung makes US and non-US versions of their SSDs? Also you said "both posters" - this thread is about the 2 review drives which both failed on Kristian Vatto.

[Hellhammer]

Originally Posted by jwilliams4200

Did you test either of the failed units with a meter to see if the power delivery circuit was providing power to the ICs?

Anand had both of the drives so I'm not sure if he did that kind of testing. He told me he's suspecting that it's the power delivery circuit but without a confirmation from Samsung, it's just a guess.

Did the second unit fail with the exact same symptoms as the first unit?

As far as I know, yes, it just died. No longer being detected by BIOS or the system at all.

What exact I/O was the second unit subjected to when it failed?

It was one of the newer enterprise tests, so not the same test as with the earlier sample. Not sure which test exactly.

Originally Posted by Coup27

Also you said "both posters" - this thread is about the 2 review drives which both failed on Kristian Vatto.

Actually, Anand killed both of the samples. I only have the regular 840, Anand has (well, had) the Pro.

[jwilliams4200]

Originally Posted by Hellhammer

It was one of the newer enterprise tests, so not the same test as with the earlier sample. Not sure which test exactly.

Can you try to find out which test was running when it died? I'd like to try it (or as close as I can replicate Anand's test with IOMeter or fio) on my 840 Pro to see if the same workload will also kill my unit.

[hhhd1]

Both drives where from the same batch that were being sent to reviewers and available to samsung officials, and both drives had the same issue.

I wonder if the same tests would be ran on a drive from a different source would it have the same issues or not.

[katkah]

Originally Posted by Coup27

Would you care to expand these comments? They suggest that Anandtech has purposely given Samsung some bad publicity at the same time as a product release.

The second statement is just thick. Do you think Samsung makes US and non-US versions of their SSDs? Also you said "both posters" - this thread is about the 2 review drives which both failed on Kristian Vatto.

Sure this site is the only one reporting a failure and at the exact same time as the release. They didn't fail last week or a week prior or even next week but within days of being released. Plextor M5P would be the next best drive NOT Crucial.

[taltamir]

Originally Posted by videoclone

Two Samsung 840 Pro review samples received by Anandtech have failed.
and they have updated the review conclusion.
*snip*
As they continue to shrink Nand, it looks like they are reducing lifespan, reducing reliability all for the sack of reducing cost. I guess the older mature 25nm/27nm NAND SSD's are the way to go.

What does NAND lifespan have to do with an SSD failing within a few days of testing?
That would be a defect in the drive. Firmware, power circuitry, controller... there are so many places where this kind of failure is far more plausible.

Now, I am actually concerned about the new intel SSD showcasing new shrunk NAND eating up lifespan when tested by anandtech. Waiting on the conclusion of the investigation they and intel are making into it

[jwilliams4200]

Originally Posted by katkah

Sure this site is the only one reporting a failure and at the exact same time as the release. They didn't fail last week or a week prior or even next week but within days of being released.

What are you talking about? You make it sound like two samples failed this week.

The first review sample failed weeks ago (in September, I believe). Then Samsung sent a replacement (I'm not sure exactly when that was received by anandtech) and that one failed, too.

Samsung really should have gotten back to anandtech with a failure report on the first unit weeks ago. I think Samsung has handled the situation very badly. Although Anand could have handled it better -- he should be printing all of the details about the two failures, such as exactly what the drives were doing when they failed, the symptoms of the failures, and any post-mortem analysis he did on the failed units.

[rgallant]

-what a mess
-the 840pro was to be the replacement for my g2 raid 0
-eta in Canada nov.16
-no firmware version listed by users or Anand that I noticed ,so I'll wait and see.
-maybe peeps with 840pro 256 can list their firmware in any comments about the 840pro 256 to compare.

[Anteaus]

Originally Posted by tweakboy

The 830 512GB came out at 1100k then price went to 700 dollars

It went from 1.1 million to 700? thats a hella crazy price drop lmao. I'm just messing with you its a funny type though.

[Coup27]

Originally Posted by katkah

Sure this site is the only one reporting a failure and at the exact same time as the release. They didn't fail last week or a week prior or even next week but within days of being released.

You obviously have no idea how stupid you sound so I think we'll just leave it there.

[hhhd1]

I really hope that samsung can get back on their feet in this problem, if they don't, the SSDs market is going to be affected badly, and price wars might not be as aggressive, specially when the 830 is being taken of the market.

[m.alkhalel]

Any 1 knows where I can find the eta to Europe ?

[katkah]

Originally Posted by jwilliams4200

What are you talking about? You make it sound like two samples failed this week.

The first review sample failed weeks ago (in September, I believe). Then Samsung sent a replacement (I'm not sure exactly when that was received by anandtech) and that one failed, too.

Samsung really should have gotten back to anandtech with a failure report on the first unit weeks ago. I think Samsung has handled the situation very badly. Although Anand could have handled it better -- he should be printing all of the details about the two failures, such as exactly what the drives were doing when they failed, the symptoms of the failures, and any post-mortem analysis he did on the failed units.

You are mouthy so I pretty much ignore you, but if it failed last month when they were tested and then sent back that should/would have been reported at the time. It wasn't reported until after the release.

[katkah]

Originally Posted by Coup27

You obviously have no idea how stupid you sound so I think we'll just leave it there.

I don't sound stupid I made a typo! The same as many other people except I was fixing it whilst YOU were sounding stupid!

[jwilliams4200]

Originally Posted by katkah

but if it failed last month when they were tested and then sent back that should/would have been reported at the time. It wasn't reported until after the release.

Why do you keep posting nonsense? Are you intentionally trying to make yourself look like you have no idea what you are talking about?

The first failure was reported in the 840 Pro review when the review came out in late September.

[Hellhammer]

Originally Posted by jwilliams4200

Can you try to find out which test was running when it died? I'd like to try it (or as close as I can replicate Anand's test with IOMeter or fio) on my 840 Pro to see if the same workload will also kill my unit.

Just spoke with Anand about this:

The first sample, like I told you before, failed while it was being filled up with 128KB sequential data (QD=1). It would no longer power on (no current detected by a multimeter).

The second sample died during power consumption test (that's either 128KB sequential data (QD=1) or 4KB random data (QD=32, 100% LBA space)). Our enterprise suite was run on it a couple of times before the failure occurred. It would still power on but was not being detected by the system/BIOS.

Ironically my regular 840 is still alive after over 40TB of writes...

[jwilliams4200]

Originally Posted by Hellhammer

Just spoke with Anand about this:

The first sample, like I told you before, failed while it was being filled up with 128KB sequential data (QD=1). It would no longer power on (no current detected by a multimeter).

The second sample died during power consumption test (that's either 128KB sequential data (QD=1) or 4KB random data (QD=32, 100% LBA space)). Our enterprise suite was run on it a couple of times before the failure occurred. It would still power on but was not being detected by the system/BIOS.

Ironically my regular 840 is still alive after over 40TB of writes...

Thanks for the additional information. I guess I will subject my 840 Pro to alternating bouts of 256GB of 128KiB QD1 sequential writes and 256GB of 4KiB QD32 random writes for a few days to see if it holds up or not.

For the 40TB-written 840, have you tried any data retention tests? For example, copy a few very large files to the SSD, compute the MD5 checksums, then power off the SSD for a week. Then power it back on and check the MD5s...

[Remobz]

Originally Posted by jwilliams4200

Thanks for the additional information. I guess I will subject my 840 Pro to alternating bouts of 256GB of 128KiB QD1 sequential writes and 256GB of 4KiB QD32 random writes for a few days to see if it holds up or not.

For the 40TB-written 840, have you tried any data retention tests? For example, copy a few very large files to the SSD, compute the MD5 checksums, then power off the SSD for a week. Then power it back on and check the MD5s...

I am crossing my fingers that all goes well.