SSDs are often "frozen" which prevents the secure erase operation taking place. To "defrost" an SSD, the usual procedure is to remove and then replace the SSD's power connector.
Assistance - Calling all Samsung EVO owners
- Thread starter Aurhinius
- Start date
SSDs are often "frozen" which prevents the secure erase operation taking place. To "defrost" an SSD, the usual procedure is to remove and then replace the SSD's power connector.
this is now the samsung bad press thread
i still think its a hdtune problem with this particular samsung drive and the reading it is getting is "buggy"
tho i have nothing to prove this and its just a guess
i would like to see any real world tests that back up hdtune's findings
if i transfer a 4gb movie i am getting 450mbps, it does not spike up and down at all, hdtune tells me i should be getting 50-100mbps!?
that to me doesnt make sense
Plenty of real world tests at the start of the thread - As I stated I first noticed the issue because an acronis true image backup of the drive was going to take 3 times longer than it used to. 30 mins + instead of 14 mins.
So real world performance issues are there Mei. You only need to initiate a file copy in windows and use res mon to watch the HDD copy speed. Mine was capping around 100mb from disk to disk.
The drive now secure erased has doubled that.
So while there was a problem with HDTune 2.55 we have shown elsewhere that not all the tests we have done are invalid and HDTune pro seems to work as does HD_Speed.
It's not about trashing samsung who have been superb. A special thanks to Memory Hanaro from the Netherlands who shipped two drives in record time (24hrs) to me and paid for collection of the old drives. Their RMA support has been superb.
Samsung the last I heard have not been able to replicate this but as we now know it seems to take some amount of time for performance to get this bad. It's obviously more than a week because I have had data on the drive for that length and others and we are not seeing the issue arise just yet.
You cannot deny an issue exists given all the people who have posted showing the issue.
We cannot deny that doing a secure erase fixes the performance problems but we all know that's exactly what a secure erase should do which is why we all did it!.
What remain's is will all our drives recreate the original performance issues over time and how long that takes.
None of us should have to secure erase a drive every 6 months. Some of the guys earlier in the thread had really new drives showing this bad 50mb/s performance already.
What we know of SSD performance is that write speed is mean't to degrade over time but read speed shouldn't degrade because reading a data block doesn't wear the nand. So the question remains why do these EVO drives get in to such a read state over time. That's something samsung need to test but it's not just one or two of us it's many people who have tested and I can bet if others did bother to test they would seem the same. I agree you wouldn't notice the issue unless you looked for it because I didn't feel any performance issues until I tried to copy files and do an acronis backup. Then I noticed and I only noticed because I remembered what it was before. If you don't watch your computer closely many issues would go unoticed.
For note I have a sandisk drive that is just as old as the samsung EVO it has had none of these issues at all and even has a pagefile on it and gets written way more. Samsung none EVO drives don't have the problems either.
So if it looks like a duck and quacks like a duck. it's a duck. If we can show a performance issue then a performance issue must exist. You might not like it and it might not bother you and that's fine but it bothers me because it's not normal and should be fixed. If my product is defective which as far as I am concerned it is because this is not normal SSD behaviour. If it happens again in 3-6months then again I'd consider my SSD defective again even though I just had two drives RMA'ed. It is not normal to have to secure erase an SSD every 3-6 months to correct dramatic read performance issues. Write performance I would accept but read performance I wouldn't. Samsung are treating this really seriously so that should tell you what we are describing isn't normal either.
Sorry if it seems like a rant but if people ignore issues they never get fixed and this should be fixed and probably can be. It's probably a bug in the garbage collection or TRIM that makes the drive read performance worse over time. That can be fixed with a firmware update probably and then none of us need to secure erase our disks. Worth the effort I think.
Last edited:
my evo is pretty new and showing 50mbps on hdtune but i just cant confirm this speed anywhere else, i have purposely not secure erased for now so i can test.
it seems every bit as fast as the pro in real world tests.
ive tried copying lots files between the 2 ssd's in windows and ive been using both with a copy of windows 8 on them and they boot up just as fast as each other
i would like to find the problem too!
i am trying to be logical about it
if you can post some file copy screenshots of low performance or there must be another way to confirm what hdtune is showing as true
everything im seeing on my drive is telling me its hdtune with the problem not the drive
but that does not explain it, and im not saying im happy with that, id like to know why and i would like more tests we can do to prove one way or the other
Take a look at the first page. John and a few others posted HD_Speed tests that show the same.
I have shown a sandisk HDTune shot and a EVO HDtune shot.
Are you suggesting that HDTune just doesn't like EVO drives and has problems benching those properly?
It's a fair suggestion but I think unlikely.
I wish I still had my old drive to show you the other tests I did but I unfortunately don't.
Download acronis free trial and set it to backup your entire EVO to another drive and check the estimated time to completion then if you have a second drive that has roughly the same amount of data try to do it back the way.
On my drives I saw a 3 fold in copy time from the EVO to the sandisk than the sandisk to the EVO.
I have shown a sandisk HDTune shot and a EVO HDtune shot.
Are you suggesting that HDTune just doesn't like EVO drives and has problems benching those properly?
It's a fair suggestion but I think unlikely.
I wish I still had my old drive to show you the other tests I did but I unfortunately don't.
Download acronis free trial and set it to backup your entire EVO to another drive and check the estimated time to completion then if you have a second drive that has roughly the same amount of data try to do it back the way.
On my drives I saw a 3 fold in copy time from the EVO to the sandisk than the sandisk to the EVO.
Okay found some -
Sandisk to EVO acronis image with res mon running
And here is the same with EVO to sandisk showing the performance halved almost.
Can't do much better than that.
One for good luck - this is the EVO doing the acronis today after being secure erased and re-imaged disk to disk so the file layout is identical
Oh and you should be looking at the true image home service that is the backup running - the top shot shows 270mb/s total where I copy from sandisk to EVO. The middle 65mb/s total shows EVO to sandisk. Bottom is the EVO today after a secure erase
Of note is the queue depth on the EVO when I copy to the sandisk middle picture jumping to 4. That's not normal. Bottom shot of the EVO today shows queue depth where it should be at 0.4 or so.
The sandisk shows less than 1 when I copy from it top shot.
Need any more convincing?
Sandisk to EVO acronis image with res mon running
And here is the same with EVO to sandisk showing the performance halved almost.
Can't do much better than that.
One for good luck - this is the EVO doing the acronis today after being secure erased and re-imaged disk to disk so the file layout is identical
Oh and you should be looking at the true image home service that is the backup running - the top shot shows 270mb/s total where I copy from sandisk to EVO. The middle 65mb/s total shows EVO to sandisk. Bottom is the EVO today after a secure erase
Of note is the queue depth on the EVO when I copy to the sandisk middle picture jumping to 4. That's not normal. Bottom shot of the EVO today shows queue depth where it should be at 0.4 or so.
The sandisk shows less than 1 when I copy from it top shot.
Need any more convincing?
Last edited:
Deleted member 138126
D
Deleted member 138126
Has anyone tested with a RAM disk? Basically:
- Copy a large file to the EVO (e.g. the 5 GB ISO I downloaded from VMware last week)
- Reboot (this will ensure none of the copied file is cached)
- After rebooting, run the RAM disk utility and present a drive large enough for the test file
- Copy the large file from the EVO to the RAM disk -- if you use Robocopy from the command line I think it will give you the duration, which will allow you to work out the throughput
From a quick Google (I've never used a RAM disk):
http://www.tekrevue.com/tip/create-10-gbs-ram-disk-windows/
By the way, I agree with Mei, I think what we are seeing is a bug between HD Tune Pro 5.5 and the EVOs. Not saying Samsung shouldn't look into it, but am not convinced it is an actual real-world performance issue. Hence my suggested RAM disk test.
- Copy a large file to the EVO (e.g. the 5 GB ISO I downloaded from VMware last week)
- Reboot (this will ensure none of the copied file is cached)
- After rebooting, run the RAM disk utility and present a drive large enough for the test file
- Copy the large file from the EVO to the RAM disk -- if you use Robocopy from the command line I think it will give you the duration, which will allow you to work out the throughput
From a quick Google (I've never used a RAM disk):
http://www.tekrevue.com/tip/create-10-gbs-ram-disk-windows/
By the way, I agree with Mei, I think what we are seeing is a bug between HD Tune Pro 5.5 and the EVOs. Not saying Samsung shouldn't look into it, but am not convinced it is an actual real-world performance issue. Hence my suggested RAM disk test.
Deleted member 138126
D
Deleted member 138126
Note, it has to be one giant file, not thousands of small ones. It is raw throughput we are looking for here.
I direct you to my res mon shots of an acronis drive image in process.
The problem isn't direct throughput copying something to the drive or off the drive from clean spare space. That is why crystal diskmark etc doesn't show the issue because that is exactly what these benchmarks do. They create a temporary file in space and read and write to it. That works and isn't the problem.
The problem comes when you are reading data off the drive that has been there for some time.
As per my drive image by acronis. As per someone's SQL tests earlier in the thread.
That is why HDTune shows the problem because it read's every bit of data off the entire drive. it doesn't create a new file.
Now I could accept that HDTune has a bug but acronis as well? Those bugs go away when you secure erase the drive?
If HDTune was reading drives badly it would likely - 1. Read all drives badly and I have shown my sandisk works fine as does other none EVO drives. 2. Not start reading the drive properly after a secure erase.
The problem isn't direct throughput copying something to the drive or off the drive from clean spare space. That is why crystal diskmark etc doesn't show the issue because that is exactly what these benchmarks do. They create a temporary file in space and read and write to it. That works and isn't the problem.
The problem comes when you are reading data off the drive that has been there for some time.
As per my drive image by acronis. As per someone's SQL tests earlier in the thread.
That is why HDTune shows the problem because it read's every bit of data off the entire drive. it doesn't create a new file.
Now I could accept that HDTune has a bug but acronis as well? Those bugs go away when you secure erase the drive?
If HDTune was reading drives badly it would likely - 1. Read all drives badly and I have shown my sandisk works fine as does other none EVO drives. 2. Not start reading the drive properly after a secure erase.
im not saying you are wrong, im saying more evidence is needed to prove it.
what works against hdtune is some on this thread have had totally different results from just switching version, that does dis-credit it a bit.
i have not seen any review sites use hdtune, rightly or wrongly.
samsung could come back and say its just hdtune, backing up what hdtune is saying with other tests allows us to say okay but then what about using this and this.
arconis screenshots are interesting but many variables there, i dont think thats enough.
just testing the evo against the pro in general use i cant tell the difference and thats what has me puzzled, the difference between 50mb and 450mb should be noticeable.
windows load times, game load times, they seem about the same.
but im not saying that is proof enough either!
lets try get to the truth then find out why
if hdtune isnt trolling us we know it affects new drives with new data too
Something is "wrong" with these Evo drives.
Yes, the earlier and later versions of HD Tune give different results, and yes, other benchmarks don't show up the issue.
What I know for sure is that before carrying out a secure erase and restore, HD Tune Pro 5.5O displayed poor results for my Evo. Also, Samsung Magician's performance benchmark was showing fairly low write IOPS.
After the erase/restore, both tests are showing good results. And so far, there has been no deterioration in those results. In fact, HD Tune results seem to have improved a little, despite me purposely adding and deleting a fair amount of data to the SSD.
I'm keeping an eye on HD Tune results. If there is any change in the next few weeks I shall report them.
Yes, the earlier and later versions of HD Tune give different results, and yes, other benchmarks don't show up the issue.
What I know for sure is that before carrying out a secure erase and restore, HD Tune Pro 5.5O displayed poor results for my Evo. Also, Samsung Magician's performance benchmark was showing fairly low write IOPS.
After the erase/restore, both tests are showing good results. And so far, there has been no deterioration in those results. In fact, HD Tune results seem to have improved a little, despite me purposely adding and deleting a fair amount of data to the SSD.
I'm keeping an eye on HD Tune results. If there is any change in the next few weeks I shall report them.
Or worth considering that none of us who have experienced these bad benchmark results have experienced any ovious drop in real world performance.
I`ve installed dozens of Evos in various PCs and laptops. Zero complaints so far.
£50 for a 120gb Evo is a good price.
Hope that helps you decide.
I`ve installed dozens of Evos in various PCs and laptops. Zero complaints so far.
£50 for a 120gb Evo is a good price.
Hope that helps you decide.
This isn't correct. I *was* experiencing poor and sporadic performance before the reset. I had to backup some data to a storage drive and was observing <30 mbps at points. That same copy now manages more than 120 mbps (other drive limitation).
It's a difficult one to narrow down but I wouldn't rubbish HDTune pro 5.5.
Yes HD Tune 2.55 which was old software anyway had a bug. Fair enough.
HD Tune pro is being used a lot now and we are showing consistent more sensible results.
Unfortunately as I have stated the nature of the problem isn't going to reveal itself with many of the other benchmarks because you are reading and writing from newly created files on to the SSD.
You need some thing that is reading existing data on the drive.
A suggestion that might show the issue -
Try to copy the contents of your OS windows\winsxs folder to another drive.
That is a large file normally and watch the estimated completion time or time it with your mobile phone.
This file is likely to be near the start of your drive as it is installed as part of your OS.
Then once copied to your other drive rename it and copy it back.
This will work if you if you have 2 SSD's with similar performance like I do.
If not, once copied back as renamed try and copy the renamed again to the second drive and compare times.
Intel's I/O bench which is much more complicate won't show this up because again it uses a method of creating a fresh file to test and that masks the issue.
In any of the drives we have shown with issues some parts of the drive are performing okay but other sections are really poor and until you try to read data out of the poor section you won't see an issue.
The reason you are probably not seeing the issues are this -
1.An SSD even with rubbish bandwidth will boot fast. It has a fast response time and windows booting generally loads lots of small files where bandwidth isn't important. Even a bad peforming SSD is going to be faster to respond than a HDD so yeah you won't notice it because it probably won't give a difference of more than 1-2 seconds which would be subjective enough for you not to notice and within a margin of error anyway.
2. The games you are likely loading are not in the bad sections of the drive and even if they are I refer you to number 1. Even with bad bandwidth it's going to appear fast unless your game is loading LARGE files in bad section of the drive.
I have already said just looking at the drive and the feel of the drive is fine. I felt the same until I investigated further with acronis and when I saw the problems with acronis I realised I needed to delve further to find the problem. It's not an easy problem to demonstrate because of the nature of it.
For exactly that reason it's why people have not been RMA'ing the SSD's left right and centre because people aren't "feeling" the problem and average user would never detect the issue or if they did they'd probably just chalk it up to a slow game or a one off.
None the less as everyone else has said - That doesn't mean there isn't an issue. There has clearly been shown to be an issue.
I wish I could show you more benchmarks but I don't have a bad drive any more as I secure erased it and it went back to samsung to test. Now I don't think they will find a fault because they won't load it up with data and use it for a month or two. It might take several weeks for the issues to start to appear. I don't think samsung will have done that test, yet.
Update -
Okay download the free trial of AIDA 64 extreme. Go to tools disk benchmark. From the drop down that says about run a linear read test on your bad drive - set the options at the top to block size 64KB to match HDTune Pro it will be set to automatic by default.
I suspect you'll get the same as HDTune pro.
Sorry I don't have a bad drive any more to test.
Yes HD Tune 2.55 which was old software anyway had a bug. Fair enough.
HD Tune pro is being used a lot now and we are showing consistent more sensible results.
Unfortunately as I have stated the nature of the problem isn't going to reveal itself with many of the other benchmarks because you are reading and writing from newly created files on to the SSD.
You need some thing that is reading existing data on the drive.
A suggestion that might show the issue -
Try to copy the contents of your OS windows\winsxs folder to another drive.
That is a large file normally and watch the estimated completion time or time it with your mobile phone.
This file is likely to be near the start of your drive as it is installed as part of your OS.
Then once copied to your other drive rename it and copy it back.
This will work if you if you have 2 SSD's with similar performance like I do.
If not, once copied back as renamed try and copy the renamed again to the second drive and compare times.
Intel's I/O bench which is much more complicate won't show this up because again it uses a method of creating a fresh file to test and that masks the issue.
In any of the drives we have shown with issues some parts of the drive are performing okay but other sections are really poor and until you try to read data out of the poor section you won't see an issue.
The reason you are probably not seeing the issues are this -
1.An SSD even with rubbish bandwidth will boot fast. It has a fast response time and windows booting generally loads lots of small files where bandwidth isn't important. Even a bad peforming SSD is going to be faster to respond than a HDD so yeah you won't notice it because it probably won't give a difference of more than 1-2 seconds which would be subjective enough for you not to notice and within a margin of error anyway.
2. The games you are likely loading are not in the bad sections of the drive and even if they are I refer you to number 1. Even with bad bandwidth it's going to appear fast unless your game is loading LARGE files in bad section of the drive.
I have already said just looking at the drive and the feel of the drive is fine. I felt the same until I investigated further with acronis and when I saw the problems with acronis I realised I needed to delve further to find the problem. It's not an easy problem to demonstrate because of the nature of it.
For exactly that reason it's why people have not been RMA'ing the SSD's left right and centre because people aren't "feeling" the problem and average user would never detect the issue or if they did they'd probably just chalk it up to a slow game or a one off.
None the less as everyone else has said - That doesn't mean there isn't an issue. There has clearly been shown to be an issue.
I wish I could show you more benchmarks but I don't have a bad drive any more as I secure erased it and it went back to samsung to test. Now I don't think they will find a fault because they won't load it up with data and use it for a month or two. It might take several weeks for the issues to start to appear. I don't think samsung will have done that test, yet.
Update -
Okay download the free trial of AIDA 64 extreme. Go to tools disk benchmark. From the drop down that says about run a linear read test on your bad drive - set the options at the top to block size 64KB to match HDTune Pro it will be set to automatic by default.
I suspect you'll get the same as HDTune pro.
Sorry I don't have a bad drive any more to test.
Last edited:
Just tested my grandfathers Evo 840 250gb, bought it a few days before mine, it's only 3 weeks old and came with the latest firmware already installed. It was a clone from a 750gb mechanical HDD.
It's showing massive drops in HD Tune 5.50. I am now have to fix this one as well.
It's showing massive drops in HD Tune 5.50. I am now have to fix this one as well.
i might give that a test before secure erase yeh
i managed to run it on pcmark8 last night while i sleep
evo
pro
the total time dont mean much, thats because one was done on 3570k and other on 4790k i think, its the individual tests times that matter and score.
it doesnt really tell me anything new, but its a long test (10 tests done 3 times each) thats meant to simulate real-world so i thought it was worth a go
it would have been nice to find out if hdtune is lying but i think im giving up on caring lol
my drive wasnt in as bad shape as some on this thread so will be pretty tough for me to prove anything with it
i managed to run it on pcmark8 last night while i sleep
evo
pro
the total time dont mean much, thats because one was done on 3570k and other on 4790k i think, its the individual tests times that matter and score.
it doesnt really tell me anything new, but its a long test (10 tests done 3 times each) thats meant to simulate real-world so i thought it was worth a go
it would have been nice to find out if hdtune is lying but i think im giving up on caring lol
my drive wasnt in as bad shape as some on this thread so will be pretty tough for me to prove anything with it
ive not seen any other ssd that has crazy dips like that
i bet i could dig out a old ocz 60gb and it wouldnt look so bad
the only *nice* theory i got is that these evo drives use a buffer, a 3gb buffer or something, and thats messing with the linear tests
but then why would that be showing up in the same place until you secure erase...
derp!
Well at least we convinced you there was an issue 
More AIDA results if you still have a bad drive will help ascert the fact there are issues here guys.
My drives are on the way to samsung but I am monitoring my new drives and as soon as I see drops like we have been seeing I'll post it again.
Your absolutely right Mei this isn't normal and everyone who has posted has shown the same thing. This should never happen to read performance on a drive.
The problem for samsung is going to be figuring out what is causing this over time. It's obviously time related because those of us who have secure erased have hammered their drives a bit and still seeing great performance which is how it should be.
More AIDA results if you still have a bad drive will help ascert the fact there are issues here guys.
My drives are on the way to samsung but I am monitoring my new drives and as soon as I see drops like we have been seeing I'll post it again.
Your absolutely right Mei this isn't normal and everyone who has posted has shown the same thing. This should never happen to read performance on a drive.
The problem for samsung is going to be figuring out what is causing this over time. It's obviously time related because those of us who have secure erased have hammered their drives a bit and still seeing great performance which is how it should be.
Deleted member 138126
D
Deleted member 138126
Using the winsxs folder as a benchmark is a terrible idea -- it contains many thousands of small files, generating tens of thousands of individual IOPS, which will bottleneck in many different places (RAM, CPU, the bus, the SATA controller, the controller on the SSD, etc.).
Now that you mention it, this could explain why your Acronis backup is so slow. If it is doing a file-level backup (as opposed to a block-level backup, which doesn't care about individual files and backs up linearly using large requests), then many thousands of small files give dramatically lower total throughput than one giant file. And you can't benchmark one SSD by copying data to another SSD -- that is 1 variable too many (hence my post on using a RAM drive).
The difference in total throughput on any drive, SSD or not, between small random-access and large linear requests is night and day. On mechanical drives there was the massive head-seek penalty that SSDs don't have, but there is still a significant difference. Look at any review benchmark and you will see that as request size increases (typically the benchmarks start at 0.5 kB and go up to 8 MB) the total throughput of the test goes up from a few 10s of MB/s to 100s of MBs/s.
Is there something funny with EVO drives, and some benchmark tools and secure erase? Yes, it certainly looks to be that way. But is there a real-world performance issue? I am not convinced.