New SSD Woes

BloodyIron

2[H]4U
Joined
Jul 11, 2005
Messages
3,439
Okay so this is the second unique SSD I've had in my hands for testing purposes.

Back near the turn of the year I had a Corsair 40GB SSD ( http://www.memoryexpress.com/Products/PID-MX30596(ME).aspx ).


It's initial speeds before putting data on it were impressive. As soon as I put data on it, it dropped to slower than my 1TB HDD. It would read at about 110MB/s or so, and no faster.

The odd thing was when I picked up a replacement I was more careful of my initial testing. I put a small amount of data on it at first (about 300MB or so), and the beginning of the drive dropped to this 110MB/s ish speed, but the rest was at the advertise >200MB/s speed. As I added more data the rest of the drive slowed down.

After further testing I never found a way to improve it and eventually returned it.



I figured I would do it again and give it the old college try with a completely different one ( http://www.memoryexpress.com/Products/PID-MX31438(ME).aspx ).

Initially before I put data on it the speeds seem much more stable than the Corsair one, pretty steady at about 230MB/s. Until I put data on it.

I dumped TF2 and BC2 onto it, and it is now down to the same kind of performance I was seeing in the Corsair.


The speeds spike up and down for the first part of the drive between 80MB/s and 125MB/s. It forms a slow drop and at the end of the drive it has dips down to as low as 40MB/s and upwards of 70MB/s.


To clarify some things, these are important specs.

WIndows 7 64bit
TRIM is running (last time I checked)
I am not using this as an OS drive, I am isolating this from it doing anything else
I have ran my tests multiple times to see if things change.


On the Corsair I used ATTO testing (which always reflected the "rated" speeds), HD Tach and HD tune. HD Tach and HD Tune showed this decline in performance, and alternative file copying testing also did too.

On the Kingston I forego the ATTO test as it seems irrelevant to my interests and tried HD Tach and HD Tune. HD Tach failed in testing (for some unknown reason) and HD Tune has shown the trends I described above. Plus the drive has appeared to have slowed down after copying large ammounts of data.




Now, are my expectations of these drives completely out of whack? Or are these things supposed to ACTUALLY GO the fucking speed they're rated?

The reason I picked up this Kingston SSD was because I read reviews in advance which showed HD Tach to have a much better performance trend, however I suspect that the people doing these tests never once actually put data on the drives.


If you know some things I can do to improve my situation I'm all ears, but if I can't figure this shit out I am returning this drive in a few days and giving up on SSDs until like say 6months from now in which I will do the same kind of testing again.

I am really turned off by these results, I must say.
 
Do you have all the latest system drivers installed?
AHCI enabled?
hibernation disabled
pagefile moved to another disk
etc etc etc
 
I ran CDM,

5x 1000MB

XXXX/Read/Write
Seq/249.8/117.0
512k/174.2/58.31
4k/12.82/19.20
4kQD32/13.21/14.25


Seems a bit short on 4k?

Also, is there another test I can use to backup these sequential results? I dont like trusting just one test.
 
Do you have all the latest system drivers installed?
AHCI enabled?
hibernation disabled
pagefile moved to another disk
etc etc etc

AFAIK latest system drivers for everything.
ACHI is enabled.
Hibernation is disabled AFAIK.
Pagefile is irrelevant as the OS is not on this disk.
 
The increase in benchmarking storage is interesting. Not that long ago almost no one would benchmark their new 320 or 500 GB HDD. Now with all the claims of SSD's being blazing fast people are benchmarking their SSD's left and right.

Does the performance of the drive feel compromised? Are you happy with the speed? Really if you want the huge numbers you're going to have to get a big RAID 0 array on either an Areca or LSI controller. I just use SSD's because they make my life smoother.
 
The increase in benchmarking storage is interesting. Not that long ago almost no one would benchmark their new 320 or 500 GB HDD. Now with all the claims of SSD's being blazing fast people are benchmarking their SSD's left and right.

Does the performance of the drive feel compromised? Are you happy with the speed? Really if you want the huge numbers you're going to have to get a big RAID 0 array on either an Areca or LSI controller. I just use SSD's because they make my life smoother.


When I pay about $2/GB I want that price to be justified. I don't buy the SSD for the storage space, I pay for it to be substantially faster than my spinning disk. As such I am trying to prove I am actually receiving what I have paid for.

I don't understand how this was not clear.
 
Back
Top