Personal computing discussed

Moderators: renee, morphine, Steel

 
thorz
Gerbil Team Leader
Posts: 214
Joined: Thu Feb 28, 2002 7:00 pm

Re: Samsung 840 & 840 EVO Bug

Fri Jun 19, 2015 1:17 am

So where do i find this new firmware?
I have an 840 EVO 750GB and the latest firmware I find on Samsung's pages is from October 2014!
Damn Samsung... they should issue a recall of this peace of junk.
 
JustAnEngineer
Gerbil God
Posts: 19671
Joined: Sat Jan 26, 2002 7:00 pm
Location: The Heart of Dixie

Re: Samsung 840 & 840 EVO Bug

Fri Jun 19, 2015 4:30 am

Install the Magician software and let it do the updating.
· R7-5800X, Liquid Freezer II 280, RoG Strix X570-E, 64GiB PC4-28800, RTX3080, 2TB SX8200Pro +2TB MX500 +NAS, Define 7 Compact, Focus PX-850, S3220DGF +32UD99, FC900R OE, DeathAdder Chroma
 
just brew it!
Administrator
Posts: 54499
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Samsung 840 & 840 EVO Bug

Fri Jun 19, 2015 5:46 am

Not sure what page you're looking at, their main SSD firmware download page (http://www.samsung.com/global/business/ ... loads.html) includes the April 15 EXT0DB6Q firmware release for the 840 EVO.
Nostalgia isn't what it used to be.
 
thorz
Gerbil Team Leader
Posts: 214
Joined: Thu Feb 28, 2002 7:00 pm

Re: Samsung 840 & 840 EVO Bug

Fri Jun 19, 2015 7:25 am

just brew it! wrote:
Not sure what page you're looking at, their main SSD firmware download page (http://www.samsung.com/global/business/ ... loads.html) includes the April 15 EXT0DB6Q firmware release for the 840 EVO.


I have looked on the US and the UK sites:
http://www.samsung.com/uk/support/model/MZ-7TE750BW
Thanks for that link.
 
Jigar
Maximum Gerbil
Posts: 4936
Joined: Tue Mar 07, 2006 4:00 pm
Contact:

Re: Samsung 840 & 840 EVO Bug

Fri Jun 19, 2015 9:32 am

OK here is the bench result of my 850 EVO 120GB drive. I am going to post results every 3 months to keep tab of how it is performing.

Image
Image
 
Jigar
Maximum Gerbil
Posts: 4936
Joined: Tue Mar 07, 2006 4:00 pm
Contact:

Re: Samsung 840 & 840 EVO Bug

Sun Aug 16, 2015 2:43 am

Follow up after almost 2 months... I was curious to see the results in Windows 10

Image
Image
 
cobalt
Gerbil First Class
Posts: 171
Joined: Mon Oct 30, 2006 11:28 am

Re: Samsung 840 & 840 EVO Bug

Wed Aug 19, 2015 10:11 pm

cobalt wrote:
And here's my "after" result. I gotta say, it's the slow access times that worry me possibly even more than the slow reads. At some point I suspect it will have degraded so far it won't be able to read that data at all. I'm happy that the access times are fixed as well.

Image

As per Chrispy's advice, set DiskFresh to launch once a month.

(And thanks, all, for the moral support and advice!)


So I've skipped my DiskFresh runs, but instead checked it out monthly with HDTune to make sure I wasn't in any immediate danger of data loss. Figured after three months, it was time to post an update:

Image

Yep, vanilla 840, clearly on the decline -- many dips below 300MB/s, and some down near 200MB/s.

Samsung: you want to claim, once again, that it only affects the 840 EVO, and not the standard 840? I guess I should just get over it, but either their QC and diagnostic capabilities are pathetic (if everyone can reproduce the problem but them), or they're lying sacks of expletive, so as it stands I'm never buying one of their drives ever again and will recommend against anyone else ever purchasing one. I'll probably continue to remind everyone about this problem any time a Samsung drive gets mentioned as well, though I'm pretty lazy, so that might work in their favor.

And to think, the only thing it would have taken for me to forgive them was an admission that the 840 suffered from the problem as well and add support for their "refresh" button to the 840.

Jerks.
 
localhostrulez
Minister of Gerbil Affairs
Posts: 2481
Joined: Sun Mar 09, 2014 11:26 pm

Re: Samsung 840 & 840 EVO Bug

Wed Aug 19, 2015 11:20 pm

Yep, I've washed my hands clean of Samsung garbage. Same with Crucial - talk about firmware bugs. Seems like Sandisk is my SSD vendor at the moment - an M500 resulted in constant crashes when I tried to run Windows Updates on a Windows 7-equipped 6530b (really, a drive that's only half stable, and passed QC?). Although I do have another 6530b with a different M500 that works all right. But meanwhile, a Sandisk Ultra II did the job just fine where the M500 barfed, and the machine feels a bit faster too. (Somehow M500's make Windows feel a tad slow for an SSD. Turns out that the issue isn't the Core 2 in that machine, it's that M500's are just sluggish.)
 
Chrispy_
Maximum Gerbil
Posts: 4670
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Samsung 840 & 840 EVO Bug

Mon Aug 24, 2015 4:02 pm

I replaced the POS 840 in my workstation about six weeks ago. That means it's unpowered, sitting on a shelf.

If I can remember I'll plug it in and test it in the new year or something to see what it's like after 6 months.
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
Arvald
Gerbil Elite
Posts: 761
Joined: Tue Sep 27, 2011 12:14 pm
Location: Gerbil-land, Canada

Re: Samsung 840 & 840 EVO Bug

Mon Aug 31, 2015 8:33 am

I ran the analysis again and my random write IOPS dropped to about 20% of what was normal. most of the other scores were dead on.

I used the "performance optimization" in Magician and after a 30 minute run I ran the tests again and all scores posted about 10% better than standard.

so the firmware did not fix issues but they gave a tool to boost yourself. only issue is only power users will find it.
 
ronch
Graphmaster Gerbil
Posts: 1142
Joined: Mon Apr 06, 2009 7:55 am

Re: Samsung 840 & 840 EVO Bug

Thu Oct 01, 2015 10:14 am

840 EVO user here. I'm using the second-to-the-latest firmware for my SSD and I couldn't flash my drive because I think there's an issue with using Magician to flash the firmware when the drive is hooked up to an AMD chipset (in my case, it's the 990FX). I tried the DOS option too but I think the CD didn't work (can't really remember what went wrong). So I can only use Magician after a reformat and these days I'm too lazy to bother with it. I think I've flashed my drive 2 or 3 times and I'm frankly a little tired of this issue. Maybe I just don't really care anymore as long as the darned thing is working.

But my question is, does the latest firmware really fix the problem? Anyone else having trouble with the DOS route and/or AMD chipsets? Some of you guys are saying the latest firmware didn't really fix the issue.
NEC V20 > AMD Am386DX-40 > AMD Am486DX2-66 > Intel Pentium-200 > Cyrix 6x86MX-PR233 > AMD K6-2/450 > AMD Athlon 800 > Intel Pentium 4 2.8C > AMD Athlon 64 X2 4800 > AMD Phenom II X3 720 > AMD FX-8350 > RYZEN?
 
Topinio
Gerbil Jedi
Posts: 1839
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Samsung 840 & 840 EVO Bug

Thu Oct 01, 2015 11:51 am

I think the answer is that the issue is unfixable, but the symptoms are prevented by rewriting the data. I don't know whether Samsung have achieved this by having it done by code in the latest firmware or the latest Magician software, but apparently they have done it.

I realise I can test this by digging out an 840 EVO from the junk pile and moving my main box's Fedora install from my Kingston drive to the POS, if it doesn't degrade the mitigation must be in the firmware itself and if it does degrade then :roll: ...
Desktop: 750W Snow Silent, X11SAT-F, E3-1270 v5, 32GB ECC, RX 5700 XT, 500GB P1 + 250GB BX100 + 250GB BX100 + 4TB 7E8, XL2730Z + L22e-20
HTPC: X-650, DH67GD, i5-2500K, 4GB, GT 1030, 250GB MX500 + 1.5TB ST1500DL003, KD-43XH9196 + KA220HQ
Laptop: MBP15,2
 
Welch
Grand Gerbil Poohbah
Posts: 3582
Joined: Thu Nov 04, 2004 5:45 pm
Location: Alaska
Contact:

Re: Samsung 840 & 840 EVO Bug

Thu Oct 01, 2015 12:26 pm

He act of writing an image to the drive itself would refresh the cells. However to answer your question, the firmware is responsible for periodically refresh the cells to avoid the read performance degradation. The software however has a faster refresh function that does the entire drive in case you don't want to wait for a specific affected area to be refreshed. If you arsn't feeling large hits in performance or you aren't in some rush then simply let the firmware do its job. Using the software to perform the "fix" actually wears down the drive faster and eats up some of your drives lifetime write cycles before it reaches end of life.
"I think there is a world market for maybe five computers."
Thomas Watson, chairman of IBM, 1943

1600x | Strix B350-F | CM 240 Lite | 16GB 3200 | RX 580 8GB | 970 EVO | Corsair 400R | Seasonic X 850 | Corsair M95 / K90 | Sennheiser PC37x
 
Chrispy_
Maximum Gerbil
Posts: 4670
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Samsung 840 & 840 EVO Bug

Thu Oct 01, 2015 12:46 pm

Ronch, just use Diskfresh.

The firmware isn't a fix, it's a workaround that's essentially indentical to scheduling a Diskfresh every few weeks. Set and Forget!
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
morphine
TR Staff
Posts: 11600
Joined: Fri Dec 27, 2002 8:51 pm
Location: Portugal (that's next to Spain)

Re: Samsung 840 & 840 EVO Bug

Thu Oct 01, 2015 1:16 pm

Arguably, just installing the firmware and letting the drive take care of it is easier :)
There is a fixed amount of intelligence on the planet, and the population keeps growing :(
 
ronch
Graphmaster Gerbil
Posts: 1142
Joined: Mon Apr 06, 2009 7:55 am

Re: Samsung 840 & 840 EVO Bug

Sat Oct 03, 2015 12:05 pm

Welch wrote:
He act of writing an image to the drive itself would refresh the cells. However to answer your question, the firmware is responsible for periodically refresh the cells to avoid the read performance degradation. The software however has a faster refresh function that does the entire drive in case you don't want to wait for a specific affected area to be refreshed. If you arsn't feeling large hits in performance or you aren't in some rush then simply let the firmware do its job. Using the software to perform the "fix" actually wears down the drive faster and eats up some of your drives lifetime write cycles before it reaches end of life.


So I suppose it's not so bad to be using the second-to-the-latest firmware, EXT0CB6Q. IIRC this firmware still didn't have the periodic refresh feature. The latest firmware is EXT0DB6Q.
NEC V20 > AMD Am386DX-40 > AMD Am486DX2-66 > Intel Pentium-200 > Cyrix 6x86MX-PR233 > AMD K6-2/450 > AMD Athlon 800 > Intel Pentium 4 2.8C > AMD Athlon 64 X2 4800 > AMD Phenom II X3 720 > AMD FX-8350 > RYZEN?
 
IAmGhostDog
Gerbil Elite
Posts: 638
Joined: Sun Oct 10, 2004 10:56 am
Location: Milwaukee, Wisconsin

Re: Samsung 840 & 840 EVO Bug

Sun Oct 04, 2015 9:32 am

morphine wrote:
Arguably, just installing the firmware and letting the drive take care of it is easier

Correct.
However. After installing latest version of Magician it wouldn't update firmware so I couldn't run Advance Performance.
What I did:
Downloaded the iso from Samsung's website.
Created a boot disk.
Booted from it.
Firmware updated.
Problem solved.
Gigabyte GA-990FXA-UD3 | AMD FX8350 Black Edition | Samsung 840 EVO 500GB | Corsair XMS3 16GB DDR3 | XFX R9 390 | Win7 x64
HT Omega Claro | Technics SL-1210 MKll w/ Stanton 681EEE MKlll | Klipsch Promedia v4.1 | Sennheiser HD 580 Precision
 
Deanjo
Graphmaster Gerbil
Posts: 1212
Joined: Tue Mar 03, 2009 11:31 am

Re: Samsung 840 & 840 EVO Bug

Sun Oct 04, 2015 9:38 am

localhostrulez wrote:
Yep, I've washed my hands clean of Samsung garbage. Same with Crucial - talk about firmware bugs. Seems like Sandisk is my SSD vendor at the moment - an M500 resulted in constant crashes when I tried to run Windows Updates on a Windows 7-equipped 6530b (really, a drive that's only half stable, and passed QC?). Although I do have another 6530b with a different M500 that works all right. But meanwhile, a Sandisk Ultra II did the job just fine where the M500 barfed, and the machine feels a bit faster too. (Somehow M500's make Windows feel a tad slow for an SSD. Turns out that the issue isn't the Core 2 in that machine, it's that M500's are just sluggish.)


Odd, I haven't had a single issue with the M500's I have (500, 250 and 120 varieties).
 
Chrispy_
Maximum Gerbil
Posts: 4670
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Samsung 840 & 840 EVO Bug

Sun Oct 04, 2015 4:05 pm

morphine wrote:
Arguably, just installing the firmware and letting the drive take care of it is easier :)

Not for Ronch, his AMD is FUBAR:

I couldn't flash my drive because I think there's an issue with using Magician to flash the firmware when the drive is hooked up to an AMD chipset

Could potentially be the AMD chipset I suppose, but I'd put money on flaky Samsung flashing software. Their track record is, uh, tarred-and-feathered.

Deanjo wrote:
localhostrulez wrote:
Yep, I've washed my hands clean of Samsung garbage. Same with Crucial - talk about firmware bugs. Seems like Sandisk is my SSD vendor at the moment - an M500 resulted in constant crashes when I tried to run Windows Updates on a Windows 7-equipped 6530b (really, a drive that's only half stable, and passed QC?). Although I do have another 6530b with a different M500 that works all right. But meanwhile, a Sandisk Ultra II did the job just fine where the M500 barfed, and the machine feels a bit faster too. (Somehow M500's make Windows feel a tad slow for an SSD. Turns out that the issue isn't the Core 2 in that machine, it's that M500's are just sluggish.)


Odd, I haven't had a single issue with the M500's I have (500, 250 and 120 varieties).

Crucial's MX100 (an entire batch of them) is, to date, the only model of SSD that has failed to work with Clonezilla (various versions), and I'm talking about a sample size of 50+ different make/models here....
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
IAmGhostDog
Gerbil Elite
Posts: 638
Joined: Sun Oct 10, 2004 10:56 am
Location: Milwaukee, Wisconsin

Re: Samsung 840 & 840 EVO Bug

Mon Oct 05, 2015 9:50 am

Chrispy_ wrote:
I couldn't flash my drive because I think there's an issue with using Magician to flash the firmware when the drive is hooked up to an AMD chipset

Could potentially be the AMD chipset I suppose

I'm running an AMD chipset.
Flash from iso worked for me.
Gigabyte GA-990FXA-UD3 | AMD FX8350 Black Edition | Samsung 840 EVO 500GB | Corsair XMS3 16GB DDR3 | XFX R9 390 | Win7 x64
HT Omega Claro | Technics SL-1210 MKll w/ Stanton 681EEE MKlll | Klipsch Promedia v4.1 | Sennheiser HD 580 Precision
 
Topinio
Gerbil Jedi
Posts: 1839
Joined: Mon Jan 12, 2015 9:28 am
Location: London

Re: Samsung 840 & 840 EVO Bug

Mon Oct 05, 2015 2:05 pm

Per the Samsung Magician v.4.5 Introduction and Installation Guide:

-------------------------------------------------
| Chipset | Driver    | Supported               |
-------------------------------------------------
| Intel   | Microsoft | Yes                     |
|         | Intel     | Yes                     |
-------------------------------------------------
| AMD     | Microsoft | Yes                     |
|         | AMD       | Yes(With Latest Driver) |
|         | NVIDIA    | Yes                     |
-------------------------------------------------

No clarity available on what the defnition of the "latest driver" might be, or for which AMD chipsets there exists a recent enough driver for it to be supported and whatever issues caused that caveat to be inserted into TFM to be non-issues ...

Also, it looks like validation on AMD systems may be rather low priority at Samsung:

Samsung Magician v.4.5 Introduction and Installation Guide wrote:

General Limitations
18) Firmware Update may fail on Samsung brand SSDs connected to AMD Controller.
Please retry using default sata ahci controller(Microsoft drivers)

DOS Utility Limitations
1) The DOS Utility is not compatible with pure SCSI or SATA NVIDIA/LSI/AMD chipset
drivers

RAPIDmode Limitations
10) Sometimes on AMD PC with AMD and ASMedia storage controllers it was found
that the IOs takes a longer time to complete. In such cases if Rapid was
enabled, it may get automatically disabled due to such IO errors. It may
display "Rapid is in inactive state". User has to reboot the PC to enable the
Rapid back.
11) If multiple iterations of Read and Write are performed, RAPIDmode may
become inactive due to system internal errors on some of the AMD /
AsMedia Controller or Driver.

though at least they do put the known issues into the manual.
Desktop: 750W Snow Silent, X11SAT-F, E3-1270 v5, 32GB ECC, RX 5700 XT, 500GB P1 + 250GB BX100 + 250GB BX100 + 4TB 7E8, XL2730Z + L22e-20
HTPC: X-650, DH67GD, i5-2500K, 4GB, GT 1030, 250GB MX500 + 1.5TB ST1500DL003, KD-43XH9196 + KA220HQ
Laptop: MBP15,2
 
Jigar
Maximum Gerbil
Posts: 4936
Joined: Tue Mar 07, 2006 4:00 pm
Contact:

Re: Samsung 840 & 840 EVO Bug

Mon Oct 05, 2015 2:31 pm

Purchased in June here are the latest results of 850 EVO 120gb.

Image
Image
 
ronch
Graphmaster Gerbil
Posts: 1142
Joined: Mon Apr 06, 2009 7:55 am

Re: Samsung 840 & 840 EVO Bug

Fri Nov 06, 2015 3:33 am

Hey guys, just curious.. Since I got my 250GB 840 EVO back in Dec. 2013, Samsung Magician gives me these benchmark scores. Are these in line with what you guys are getting? Other pertinent hardware: FX-8350, Gigabyte 990FXA-UD3 rev. 4 (this is my second AM3+ mobo; my first board, an MSI 990FXA-GD65, gave me the same bench scores as well), 2 x G.skill 4GB DDR3-1866, Hitachi 1TB HDD, an LG DVD-RW drive. Been using different versions of their chipset drives, and right now I'm using their latest. Windows 7 64-bit Ultimate. BIOS set to AHCI, not IDE or RAID. Firmware is EXT0CB6Q. Latest is EXT0DB6Q but I'm not sure I wanna flash it since it just rewrites data on the SSD automatically and may potentially reduce the SSD's life. I use DiskFresh instead. Obviously, when I got this SSD the firmware in it was older so I obviously got the same numbers using those older firmware versions.

Image
NEC V20 > AMD Am386DX-40 > AMD Am486DX2-66 > Intel Pentium-200 > Cyrix 6x86MX-PR233 > AMD K6-2/450 > AMD Athlon 800 > Intel Pentium 4 2.8C > AMD Athlon 64 X2 4800 > AMD Phenom II X3 720 > AMD FX-8350 > RYZEN?
 
Chrispy_
Maximum Gerbil
Posts: 4670
Joined: Fri Apr 09, 2004 3:49 pm
Location: Europe, most frequently London.

Re: Samsung 840 & 840 EVO Bug

Fri Nov 06, 2015 9:13 am

If you read back through the thread you'll see that the built-in benchmark is worthless, people posting the Samsung results giving the all-clear, whilst actual full-drive scans from HDTune or HDTach showing horrific, slower-than-5400rpm-drives transfer rates with huge access times.

If you're running Magician you're probably on the latest kludged firmware that periodically refreshes the disk in the background, so you should be okay. Run an HDTune or HDTach scan on the drive and if you get a reasonably flat line at several hundred MB/s then you're fine to ignore it and keep on trucking.

If not you can either manually force a firmware update or you can just use Diskfresh which does pretty much the same thing.
Congratulations, you've noticed that this year's signature is based on outdated internet memes; CLICK HERE NOW to experience this unforgettable phenomenon. This sentence is just filler and as irrelevant as my signature.
 
cobalt
Gerbil First Class
Posts: 171
Joined: Mon Oct 30, 2006 11:28 am

Re: Samsung 840 & 840 EVO Bug

Fri Nov 06, 2015 10:31 am

Chrispy_ wrote:
If you read back through the thread you'll see that the built-in benchmark is worthless, people posting the Samsung results giving the all-clear, whilst actual full-drive scans from HDTune or HDTach showing horrific, slower-than-5400rpm-drives transfer rates with huge access times.


I have almost nothing to add to this, except to quote and reiterate because it's such an important point:

The Samsung benchmark is meaningless -- it reports the FASTEST time it finds. Even a terribly broken 840 in use will have some spot which was written to recently and will report a fast transfer time, even if 99% of the drive is <10MB/sec. Or perhaps the Samsung benchmark is writing data and then reading it, which sidesteps the issue entirely because it's not showing you how the drive will behave with YOUR data.
 
Hance
Darth Gerbil
Posts: 7775
Joined: Mon Sep 29, 2003 1:58 pm
Location: Grace Idaho
Contact:

Re: Samsung 840 & 840 EVO Bug

Fri Nov 06, 2015 11:55 am

Just ran across this thread, my transfer rate ranges from 7 MB/s minimum to 1028 MB/s maximum with an average of 816 :-? Going to run HD Fresh and see if it gets rid of the 7 Mb/s crap.

ImageHD Tune by Hance1976, on Flickr
 
just brew it!
Administrator
Posts: 54499
Joined: Tue Aug 20, 2002 10:51 pm
Location: Somewhere, having a beer

Re: Samsung 840 & 840 EVO Bug

Fri Nov 06, 2015 12:13 pm

Hance, which Samsung do you have (840 or 840 EVO), and what firmware version are you running on it?
Nostalgia isn't what it used to be.
 
Hance
Darth Gerbil
Posts: 7775
Joined: Mon Sep 29, 2003 1:58 pm
Location: Grace Idaho
Contact:

Re: Samsung 840 & 840 EVO Bug

Fri Nov 06, 2015 1:01 pm

just brew it! wrote:
Hance, which Samsung do you have (840 or 840 EVO), and what firmware version are you running on it?



2X 840 Evo purchased right at 2 years ago. Still running whatever firmware they came with. Was going to update them but got no love from the Samsung Magician software because they are running in Raid 0 and it doesn't support raid :roll: Link to the exact drives I have from my Newegg order. http://www.newegg.com/Product/Product.a ... 6820147247

edit: I'm nervous about updating firmware where they are in raid. Don't really want to risk nuking my windows install. Getting everything setup and working on the raid controller was a way bigger PITA than it should have been.
 
sustainednotburst
Gerbil In Training
Posts: 8
Joined: Fri Jul 10, 2015 9:38 am

Re: Samsung 840 & 840 EVO Bug

Fri Nov 06, 2015 1:08 pm

ronch wrote:
Hey guys, just curious.. Since I got my 250GB 840 EVO back in Dec. 2013, Samsung Magician gives me these benchmark scores. Are these in line with what you guys are getting? Other pertinent hardware: FX-8350, Gigabyte 990FXA-UD3 rev. 4 (this is my second AM3+ mobo; my first board, an MSI 990FXA-GD65, gave me the same bench scores as well), 2 x G.skill 4GB DDR3-1866, Hitachi 1TB HDD, an LG DVD-RW drive. Been using different versions of their chipset drives, and right now I'm using their latest. Windows 7 64-bit Ultimate. BIOS set to AHCI, not IDE or RAID. Firmware is EXT0CB6Q. Latest is EXT0DB6Q but I'm not sure I wanna flash it since it just rewrites data on the SSD automatically and may potentially reduce the SSD's life. I use DiskFresh instead. Obviously, when I got this SSD the firmware in it was older so I obviously got the same numbers using those older firmware versions.

Image



The new FW is not constantly rewriting data on the SSD. The FW fix itself was a new Read code or algorithm, because the speeds come right back up to normal immediately after doing the FW update. Without any time for any refresh or rewriting to take place. The Idle refresh was an added feature as a just in case.

Allyn Malventano, reviewed the new FW on PCPER when it came out, posted some additional comments about the fix on Overclock.
http://www.overclock.net/t/1507897/sams ... t_23909666

some excerpts:

"There's no way the issue could have been mitigated without having a correct read algorithm in the first place, especially while I was running a full sequential read pass immediately after the update. It is simply not physically possible. Either they could read stale data fast or slow. The result after updating was fast. There was no forced optimization pass during the update this time around, so nothing to mask the results, either."

"5 samples that had slowed significantly did not initiate an optimization pass on their own and saw near full read speed simply by applying the update."

"Let's assume conservatively that a worst-case 840 EVO self-triggered an optimization every two months. and that the SSD was completely full during those actions (meaning the entire SSD flash space must be refreshed). That is only 6 full drive writes per year. Using an (again worst case) 1000 flash erase cycle figure, that comes to 0.6% of the drive life consumed per year. "

"Based on these results, it is unlikely that we will see patched 840 EVOs burning through flash cycles to any noticeable degree."

"Their (Samsung's) initial statement made it look like Magician was going to be a band-aid and would basically do a refresh pass every so often. This was not the actual case once we got the update in-hand and could see it in action.”
 
Hance
Darth Gerbil
Posts: 7775
Joined: Mon Sep 29, 2003 1:58 pm
Location: Grace Idaho
Contact:

Re: Samsung 840 & 840 EVO Bug

Fri Nov 06, 2015 1:16 pm

This is after running DiskFresh on them. Still lots of peaks and valleys but the overall average performance is way higher. Minimum of 700 MB/s now instead of the 7 MB/s I had before :o

ImageHD Tune 2 by Hance1976, on Flickr

Who is online

Users browsing this forum: No registered users and 1 guest
GZIP: On