Read more.The company promises another firmware update to fix the SSD's performance issues.
Read more.The company promises another firmware update to fix the SSD's performance issues.
Phew, and I just ordered an 840 last week but it went end of life so ended up with an 850 instead apparently these will have no issues?
OS and Apps on my desktop are on 840EVO's (120GB and 500GB respectively) and I've not noticed any slowdown. That said, I've not patched the firmware (from the factory supplied one) because last time I did the Samsung software came up with a message that upgrading the firmware would erase the drive, and I didn't fancy the hassle of being forced to test my backup recoveries.
Anyone going to wonder out loud why this comes to light when Samsung are doing a push on the 850's? That said, I've got 840's in the desktop and 830's in the two laptops and all four are working very well.
Slightly shocking.
I've even seen someone apply the last "fix".....it took ages to re-write every cell on the disk, a recipe for disaster.
Until the 840 EVO I wad been content that Samsung were a quality SSD supplier....I now get the feeling they are more like Corsair: Make a good product that sells extremely well and gets high praise and then release a new version with every conceivable corner cut.
Main PC: Asus Rampage IV Extreme / 3960X@4.5GHz / Antec H1200 Pro / 32GB DDR3-1866 Quad Channel / Sapphire Fury X / Areca 1680 / 850W EVGA SuperNOVA Gold 2 / Corsair 600T / 2x Dell 3007 / 4 x 250GB SSD + 2 x 80GB SSD / 4 x 1TB HDD (RAID 10) / Windows 10 Pro, Yosemite & Ubuntu
HTPC: AsRock Z77 Pro 4 / 3770K@4.2GHz / 24GB / GTX 1080 / SST-LC20 / Antec TP-550 / Hisense 65k5510 4K TV / HTC Vive / 2 x 240GB SSD + 12TB HDD Space / Race Seat / Logitech G29 / Win 10 Pro
HTPC2: Asus AM1I-A / 5150 / 4GB / Corsair Force 3 240GB / Silverstone SST-ML05B + ST30SF / Samsung UE60H6200 TV / Windows 10 Pro
Spare/Loaner: Gigabyte EX58-UD5 / i950 / 12GB / HD7870 / Corsair 300R / Silverpower 700W modular
NAS 1: HP N40L / 12GB ECC RAM / 2 x 3TB Arrays || NAS 2: Dell PowerEdge T110 II / 24GB ECC RAM / 2 x 3TB Hybrid arrays || Network:Buffalo WZR-1166DHP w/DD-WRT + HP ProCurve 1800-24G
Laptop: Dell Precision 5510 Printer: HP CP1515n || Phone: Huawei P30 || Other: Samsung Galaxy Tab 4 Pro 10.1 CM14 / Playstation 4 + G29 + 2TB Hybrid drive
I've got an 840 EVO and haven't noticed any slowdown, although apart from booting into OS I don't do much shifting of data to/from the drive.
Might look at the fix anyway, but if it formats the drive in the process I'll steer clear.
Same here, my 14/15 month old 840 Evo still runs fine, but then it's only used as the OS drive, so it's not really worked that hard compared to my other SSD's. Maybe not ALL drive are affected?
I bought my 250GB EVO on Black friday for less than 100 bucks, so I'm not complaining. I was going to wait for the 850 EVO, or even snag the pro, but the 840 came with Far Cry 4 I haven't noticed any slowdowns and I have a Y50--it's the main drive so it does all my read-writing. However, I've only owned it for a few months. I'll probably apply the new fix, but most people are noticing the slowdowns after at least a year. Either way, I still will probably never buy an SSD from anyone but Samsung. Their customer service is great and just the fact that they're trying to fix a problem with a 3-year-old SSD speaks volumes. Once I build my desktop rig I'll probably get an 850 just to be safe...
The initial fix won't wipe the drive (and is fairly fast if there's not a huge amount of data on the drive - I think around 10 minutes for 40GB when I did it a few months back) but the problem is that that initial fix (which was promised to be permanent) only has temporary benefit (couple of months) before rates start to deteriorate again.
Leety - the issue specifically affects OLD data - i.e. data that has been sitting on the drive for a long time (e.g. Windows files). Ironically, those who are frequently writing data are the ones who don't have to worry (or at least not worry as much). But certainly it's true that not all drives are affected, and the amount of wear and tear prior to applying the initial fix could possibly be a factor in determining which drives are continuing to slow down. Others have suggested temperatures but Samsung have stated that it's not temperature-related.
plexabit - It's actually less than two years old, still well within warranty... and the problem is appearing a couple of months after applying the initial fix (the fix was only issued a few months ago and already performance has started to deteriorate for some).
Hopefully not, but who can really say? For instance, Samsung themselves suggested that the first firmware update would fix the issue, by re-writing the drive and refreshing the NAND cell states and by making changes to the wear-leveling algorithms for the TLC NAND. Except that seems to have been a temporary fix. Either that, or their revised algorithms aren't any good.
This begs the question, if they don't truly understand the issue or how to fix it, what is to stop the issue occurring with their other products that use the same technology, i.e. any Samsung SSD that uses TLC NAND? Personally I find this a bit concerning, because it suggests that they either not entirely sure how to fix or, or actually the problem lies within the product design and cannot truly be fixed, in which case they'll be looking at some major class action lawsuits against them. Arguably as a product can they really vouch for their marketing claim of the drive, one of which is on their website now that says "Unmatched reliability for consistently high performance"? Erm, don't think so.
spl (23-02-2015)
I have a 830 SSD now working well with both XP and Win10. Next I'd better target the 850?
My mums 840 evo still seems to be running well after 6 months or so. I noticed yesterday that a new firmware was available for my 850 pro but after reading some of the horror stories online I think i'll just leave it on the old firmware.
I'd like to bet that there's a lot of Samsung Storage engineers looking at this problem. Yes, I know that the 850's out and Samsung's usual attitude is "throw it away and move on" (c.f. their smartphones), but the damage to reputation is something that they'd definitely want to avoid, and besides, the 840 is still available.
Personally I'd be content that IF this is a base-level issue then Samsung 'fess up and perhaps offer some kind of cash-back deal on an 850. Do that and I'd suggest that they'd not get the kind of vitriol that NVidia is currently suffering from.
What I'd definitely not want though is a class action lawsuit - that will solve exactly nothing. All a CAL will do is line the pockets of some overpriced lawyers in the USA and certainly won't get us non-US citizens a fix or some kind of redress.
Like I said - the first, (and only), time I tried to apply new firmware I got a big warning that the drive WOULD be erased. That said this was some considerable time ago (a year?) so maybe technology has moved on from that. I'll bow to your (obviously) better experience...
Hmm, I'd be very interested to learn if there's any facts* to back up what you're saying above (no insult intended), got any links? Reason I'm interested is that my two 840's are on OS and Apps respectively, so probably containing a high proportion of written-once (aka Old) data. In which case I should be seeing a dramatic fall off in response. However, (and I'm not disagreeing with you btw), those drives are also in front of the front intake fan in my case, so kept very cool. Which scant "evidence" suggests that - like Syphadeus suggests - perhaps Samsung haven't got the whole picture in front of them?
(* I'm making a distinction between proper tested explanations = facts, as opposed to the kind of "well it works for me" type stuff ... like the temperature-related thing I gave above).
Anyway folks, (and folkesses if there's any out there too), so what's the opinion of the body corporate - is this latest firmware worth applying or not? I'm getting a bit of a "if it ain't broke..." type vibe. And other than an increase in BSODs, I've not noticed any issues with my desktop, so there's a strong temptation to just leave it be.
Crossy, take a look on Techreport - they've done a better job than most sites at keeping on top of this. They did some testing with 840 EVOs owned by the TR staff and have summarised a lot of what's going on in the overclock.net forum threads about this (which is where it was first discovered, a few months before it became widely known via the tech media sites). Presumably you're not talking about facts to back up it affecting old data? That's been well-established right from the beginning.
I've had Borderlands on that drive since the get-go so that's the data I used to test (since there's some nice big files there for sequential transfer testing, whereas Windows of course is mostly small files). No apparent performance loss, so looks like I was lucky. Possibly the reason for that is that I got the fix / update installed pronto so the drive hadn't had any significant use prior to the fix.
So I'd recommend installing it (but don't be cross with me if it goes up in flames ;-)). Re testing by the way, performance loss is drastic enough that boot times become noticeably longer (we're talking drops into the low 30MB/s range in the worst cases) so it's something you'd likely spot without needing large old files. But they are nice to have for confirmation.
crossy (24-02-2015)
Oh and download the 'performance restoration' file - if you just go straight for a firmware update, that possibly would wipe the drive. But what's included in the performance restoration tool won't do.
Sound advice, and thanks for pointing me at Techreport - so many sites, so little time to browse them.
I've got Crystal benchmarks (somewhere) from when they went in - maybe time to retest and see if I'm in the "bad" area for performance. I'd also kind of assumed (wrongly from what you've said) that the perf fix was a new firmware revision rather than something distinct. Nice to have that cleared up by someone more knowledgable!
There are currently 1 users browsing this thread. (0 members and 1 guests)