Assistance - Calling all Samsung EVO owners
- Thread starter Aurhinius
- Start date
The drive got a pretty good trashing with shadowplay cache files before they changed the way the shadowplay loop was recorded (about 8tb of writes over a couple of weeks).
No problems in actual use and I can copy files from one SSD to another from what I'm pretty sure are the really low bits in that benchmark and see expected speeds, just seems to be low level disc tools that are affected.
EDIT: Tried changing link power management for the SATA interface between enabled/disabled which had no effect on the benchmark nature.
Last edited:
Nope but I've probably hammered it a lot harder than most who have owned a drive over the same time period (more than 8tb of writes on it total) - which depending on what the problem is may be why mine is so much worse than most.
I dropped a quick e-mail to Tomshardware to see if they were interested in getting involved to see if they can replicate the issues and investigate. It's such a popular consumer drive and with more folks coming forward showing problems I think we need to go wider. It'll likely come to nothing but worth a shot.
im doing some tests now
i see a problem too but only using hdtune
i cant seem to get it to show in another program
but i see others have
i couldnt try hdtach it said i needed win95 or something crazy?
and cant see a download link for atto3?
atto2.4 its not showing a problem
ill post some screenshots after i uploaded them comparing evo with pro
i guess hdtune does something very different in benchmarks
Deleted member 138126
D
Deleted member 138126
Okay not a huge surprise here.
I secure erased the OLD faulty drive to go back to Samsung and benchmarked it without even a partition installed or a format.
Not surprised at all. I will be annoyed if Samsung come back and say "we can't find a fault"
They are still saying they need this old drive to replicate the problem! I don't get how they can't recreate this given all the evidence for a problem here on the forum but maybe they aren't waiting long enough. I don't know.
Tested new drive again today - same as 2 days ago but I haven't copied any more data to the drive.
Don't send it back empty! Create a bunch of directories, and copy something big (and not personal) like a game into each directory so that it has at least 20-50 GB used. Then run the benchmark and screenshot. Even copy the installer for the exact benchmark program you are using to the root with a readme explaining this is what you've used.
it looks that way
or maybe more how some programs benchmark the drive
i dont understand how these tests work well enough
hdtune and programs that bench the same way hate the evo!?
Ok, very strange they can't reproduce. Hopefully they can on your drive. As someone suggested create some random data on it before sending with something like this.
Before samsung got involved I did consider contacting review sites, it's worth a try. I'll link this thread on storagereview, they did a review but didn't uncover this, hopefully they still have an EVO.
Possibly but none benchmark situations have been shown to be effected aswell.
Last edited:
hm yeh but it does seem to react very badly to files already written on the evo, where as the pro there is no difference where files are written or not
its hard to believe that 50mbps hdtune shows is accurate when i copy a video file from one to the other i cant get a screenshot because it happens so fast
i wanted try pcmark8 as thats what hexus use in their consistency test but you have to pay for that part of it
my gut tells me its just a bug with evo and that test but it should be looked into with some better software than i have
maybe we can get hexus to update their benchmarks, they never did that part of the test with the evo that i see
This is how things are looking after making an image of my SDD, performing a secure erase, then restoring from the image.....
Being a bit of a numpty, I had done a "before" screenshot, but cleverly saved it to the SSD after I'd made the image. However, the before was very much like this.....
The secure erase and restore has made a massive difference. I'll test again after a few days use. In fact, I'll shift a few files around and see what happens.
Being a bit of a numpty, I had done a "before" screenshot, but cleverly saved it to the SSD after I'd made the image. However, the before was very much like this.....
The secure erase and restore has made a massive difference. I'll test again after a few days use. In fact, I'll shift a few files around and see what happens.
It's on an Intel SATA 3 port.
The "good" graph is after I carried out a secure erase and restored everything from an image I made before the secure erase.
I used EaseUs Todo Backup to carry out the backup/restore. The restore was carried out with "Optimize for SSD" enabled.
The "good" graph is after I carried out a secure erase and restored everything from an image I made before the secure erase.
I used EaseUs Todo Backup to carry out the backup/restore. The restore was carried out with "Optimize for SSD" enabled.
That's what I was thinking.
The graph looks much more "healthy" after the secure erase. The performance of my machine doesn't seem to have improved though (I`ve always thought that the transfer rate of SSDs isn't the main reason they're fast, it's the low access times).
I'm going to shuffle around some files (large and small) and see if the graph starts to get messy again.