Samsung’s fix for the EVO tested!
** Edit **
The tool is now available for download from Samsung here. Another note is that they intend to release an ISO / DOS version of the tool at the end of the month (for Lunix and Mac users). We assume this would be a file system agnostic version of the tool, which would either update all flash or wipe the drive. We suspect it would be the former.
** End edit **
As some of you may have been tracking, there was an issue with Samsung 840 EVO SSDs where ‘stale’ data (data which had not been touched for some period of time after writing it) saw slower read speeds as time since written extended beyond a period of weeks or months. The rough effect was that the read speed of old data would begin to slow roughly one month after written, and after a few more months would eventually reach a speed of ~50-100 MB/sec, varying slightly with room temperature. Speeds would plateau at this low figure, and more importantly, even at this slow speed, no users reported lost data while this effect was taking place.
An example of file read speeds slowing relative to file age.
Since we first published on this, we have been coordinating with Samsung to learn the root causes of this issue, how they will be fixed, and we have most recently been testing a pre-release version of the fix for this issue. First let's look at the newest statement from Samsung:
Because of an error in the flash management software algorithm in the 840 EVO, a drop in performance occurs on data stored for a long period of time AND has been written only once. SSDs usually calibrate changes in the statuses of cells over time via the flash management software algorithm. Due to the error in the software algorithm, the 840 EVO performed read-retry processes aggressively, resulting in a drop in overall read performance. This only occurs if the data was kept in its initial cell without changing, and there are no symptoms of reduced read performance if the data was subsequently migrated from those cells or overwritten. In other words, as the SSD is used more and more over time, the performance decrease disappears naturally. For those who want to solve the issue quickly, this software restores the read performance by rewriting the old data. The time taken to complete the procedure depends on the amount of data stored.
This partially confirms my initial theory in that the slow down was related to cell voltage drift over time. Here's what that looks like:
As you can see above, cell voltages will shift to the left over time. The above example is for MLC. TLC in the EVO will have not 4 but 8 divisions, meaning even smaller voltage shifts might cause the apparent flipping of bits when a read is attempted. An important point here is that all flash does this – the key is to correct for it, and that correction is what was not happening with the EVO. The correction is quite simple really. If the controller sees errors during reading, it follows a procedure that in part adapts to and adjusts for cell drift by adjusting the voltage thresholds for how the bits are interpreted. With the thresholds adapted properly, the SSD can then read at full speed and without the need for error correction. This process was broken in the EVO, and that adaptation was not taking place, forcing the controller to perform error correction on *all* data once those voltages had drifted near their default thresholds. This slowed the read speed tremendously. Below is a worst case example:
We are happy to say that there is a fix, and while it won't be public until
some time tomorrow now, we have been green lighted by Samsung to publish our findings.
I tried to use the windows
I tried to use the windows program to no avail. The problem with the AMD driver was unsolvable to me. I used the DOS tool and it worked. But it has only has two steps. When i run it again, the windows tool continued to work. It blazed through steps 1 and 2 and did a long step 3.
Like someone said here, it´s absurd that the Samsung Magician did not has given any notice about the new firmware or about this very grave problem. I got to know it for mere luck listening to TWiT.
I noticed my system not so fast for SSD for some time but blamed Windows. My Steam’s The Witcher 2 install is in the SSD and the load times were terrible. Inital tests showed a perceivable improvement.
Thanks for PCper for the heads up about this grievous problem.
Thanks for sharing your
Thanks for sharing your findings with the Samsung Performance Restoration Tool for 840 EVO.
Have there been any tests with an 840 EVO that is encrypted with bitlocker? (Testing after unlocking, and testing with files locked)
Yes, I have updated my Mac
Yes, I have updated my Mac with a CD burned from the ISO provided by Samsung. But for the downloaded Samsung ISO to boot a Mac from a CD, you have to change it to a DMG (renaming it is enough); at that point you can burn the image to CD from Disk Utility and it will boot and run. I documented the steps here:
As far as I can tell the Samsung USB key software seems to be impossible to use on a Mac as downloaded, because it requires a proper EFI. That means you have to assemble the bootable USB key yourself using software from multiple sources on the net, and using the Terminal is practically required. Naturally, all of that is a bit of a challenge except for advanced users. Since my Mac still has an optical drive, I decided to burn a CD-RW instead because that took 5 minutes. (You can boot an Intel Mac from DOS on a CD; that doesn’t have the same EFI restriction as booting from USB keys.)
Overall the Samsung software for the Mac is too much of a challenge for less technical users because the Samsung software is incomplete (not ready-to-run as downloaded), yet the Samsung documentation, which is written for IT types, doesn’t fully disclose how incomplete it is.
Thank you for these excellent
Thank you for these excellent posts. I have a Macbook Pro with a Superdrive so I was able to create a CD using the .iso file copied to a .dmg then used Disk Utility to burn. I have both OSX and Bootcamp (Win 7.1) partitions on the 840EVO 512G. Could not find any online references to anyone who had OSX and Windows on the same disk so I decided to strike out on my own.
0. Made sure my OSX volume had recently been backed up via Time Machine.
1. Downloaded the .iso file in OSX, copied to .dmg, then burned CD using Disk Utility. Thanks, Conrad, for the excellent procedure.
2. Rebooted into Bootcamp, downloaded the .zip file for Windows, extracted the contents, and ran the executable. It reboots the system once back into Windows. Firmware got updated and the NTFS partition was processed in three stages. I have about 100GB formatted as NTFS and it took less than an hour. Documentation states the Windows version will work only on NTFS, so I decided I’d better also run the bootable CD to do the HFS OSX partition.
3. Rebooted into Windows, no problems.
4. Restarted from OSX, holding down option key until presented with multiple boot disk options, selected the CD.
5. CD booted into DOS, did not update firmware, then started Stage 1 which went pretty fast. Stage 2 was chugging along then my screen went blank and nothing seemed to be happening. So I shut off the computer and restarted, holding the option key again until I could boot from the CD again. Firmware skipped, Stage 1 done fast, Stage 2 began again. I left it alone and went to bed.
6. Next morning, Stage 2 was complete, taking 4 hours and 34 minutes. Back to a DOS prompt. Wow, so 1980’s. No command to reboot here (put in a ? and got a list of commands, and EXIT did nothing). Powered off.
7. Held the option key down and restarted. Selected the OSX partition and booted up. All good.
I ran no before or after benchmarks but subjectively programs seem to load a bit faster.
Any idea when these drives
Any idea when these drives rolling off the assembly line will have this fix already incorporated in them? I bought one for my daughter a few weeks ago, and a friend ordered one a few days ago. Is there a serial number we can go by?
Allyn, is there any update on
Allyn, is there any update on the performance of the “stale” drives that had been touched with the performance tool? I know you were holding on to for testing purposes to see if the tool did what it was supposed to do and didn’t have any adverse effects that would creep back up.
I’ am still getting half
I’ am still getting half speeds of the reads on my 840’s.. anyone else? From 1100 down to 500 in raid 0.
I did the updates prior to
I did the updates prior to raiding the drives. FYI.
This is a recent event as
This is a recent event as well guys. I don’t think that the flaw is really fixed.
It may be in a non-raid
It may be in a non-raid configuration…I’m just saying..Al help me out here please after the Holidays! and by the way…. I really want to win that car!
I am going to replace my CD
I am going to replace my CD SuperDrive with a 840 EVO drive. I have thunderbolt on my iMac. I have an Kannex Thunderbolt to USB 3 and esata connectors. Can I use the esata connector for this updating? I will power thru another source.
Should I run this software if
Should I run this software if I’m having no issues or wait and use it if I ever see performance drop?
My SSD EVO 840 500GB is
My SSD EVO 840 500GB is having a very bad performance.
– Sequential Read: 68MB/s – max 540MB/s
– Sequential Write: 62MB/s – max 520Mb/s
– Random Read : 14973 – max 98000
– Random Write : 10862 – max 90000
I have the latest Magician software downloaded (v46)
Magician also updated the firmware to version : EXTODB6Q, which is the latest. However, the performance is still the same.
I’ve heard and read about Performance Restoration tool from Samsung, but could not find in Samsung download site.
Can you please let me know where I can find this software? What are the other options?
I had some bad luck following
I had some bad luck following the suggestions ubove for updating the firmware on my RAIDed Evo 840 and feel I should write this to caution others. Take note that I did have a BSOD after I successfully de-RAIDed my 2xEvo 840s in Windows 7, but not right away – a few minutes after while on the desktop.
I have an intel comtroller with 6 ports and two RAIDs on it:
1) 2xSamsung Evo 840 256gb in RAID0
2) 4xSeagate SSHD 2tb in RAID10
I followed all the instructions but when I completed step no. 7 and re-enabled RAID in BIOS – my other 4xSeagate SSHD 2tb RAID had ‘FAILED’. I am still in the stage of fixing it but I felt I had to come up here and give some warning so that other could benefit if they had this happen to them.
If this happens to you – you can still salvage your data from the ‘FAILED’ RAID. What I did was:
1) Most important thing is not write or initialize or do anything destructive to any of the disks from the lost array.
2) Go to intel RAID in bios and ‘Reset Disks to Non-RAID’ and choose all of the disks from the failed array.
3) In windows do nothing to the separate disks (if disk manager pops up telling you to initialize them or something).
4) Get an empty disk big enough to hold all your data from the array (I had to buy one 4tb disk)
5) Download “ReclaiMe Free RAID Recovery” and use it to find the original RAID configuration from the now-separate disks and rebuild it (you could use any other software if you like).
6) Write the RAID from the failed now-separate disks to the empty disk you have set up.
7) Redo the ‘FAILED’ RAID from scratch and put your data back on it.
I still didn’t do those two last parts as I am waiting for the 4tb disk I bought to finish formatting – so I hope it actually IS free. I have used reclaiMe to see if my data was actually still there and I was very relieved to find out that it was all still there.
There are other things you could do and I have tried using testdisk to write the partition table and when that didn’t fix my problem I tried copying the backup boot sector with it but my backup was bad (maybe had nothing to do with my problem as it was not used as an OS volume – but tried that anyway). When testdisk didn’t fix my problem I defaulted to use the solution stated above.
I hope this helps anyone who found himself/herself in the same situation.. and good luck.
If this is a common problem when you have more than one RAID on the controller it would be a really good idea to add some warning in the article itself.
well…here it is, september
well…here it is, september 2015, and i still cannot get my DTX06B0G firmware upgraded…i have magician v.7 and still nothing…error message “firmware upgrade failed”…
c’mon guys, what is going on!…cannot get to samsung to ask why, no FAQ’s on the problem (except to refer to magician, which has NEVER worked for the upgrade), my big bucks ssd has just been tested to be SLOWER than my WD platter!…and still no resolution from samsung!
i will not recommend samsung for ssd’s, due to their incompetence, indifference, and aloofness (no contact info)…cannot contact them for advice, the software just does not work, and any faq’s point back to the worthless magician software firmware upgrade…and for this i paid big bucks?