If you partition the the Corsair drive the same way you have with the Torqx 2, you get 6 to 20 times better performance? I see you partition the drive 3 times? You also understand the more capacity of the drive is filled the lower the performance you will see. Also after all those writes you do to the drive and then benchmarking them right after can affect the performance results as well. I have to say you are using the drive under different circumstances and also you are comparing a 32GB to a 60GB.
Can you also check what firmware version you are running? You can use a tool called Crystal Disk Info to check this information.
Honestly, with what you are saying - i.e. that it is supposedly "normal" that the Torqx is 6 to 20 times slower than a mainstream Corsair drive when it has two (not three!) partitions and is half full:
-- either (sorry) you are bullshitting me (--> meaning that this miserable block writing performance is NOT normal) which then in turn means that the drive is must be broken.
-- or it is true (with Patriot drives, that is?)
In both cases, the result is the same however: I won't waste any more time with this bad drive because it either IS terrribly slow or it is simply broken.
I bought another Corsair yesterday, so this is moot anyway.
For completeness, below's the result of Crystal Disk Info.
It could be a defective drive. I am having our team look into your issue and try to set up the same set up as you to try and replicate the performance you are getting. We can go ahead and get that drive replaced and if you wish not to use the Torqx2 drive anymore I can understand but please do not talk down on me. I do not see how that helps anyone.
I'm not talking down on you. My point was that you were implying in your post #11 that in my setup circumstances (partitions, usage, free space) obviously it COULD BE that I get this "6 to 20 times" lower block write speed.
You are on the side of the company, so you must know the properties of your drives, thus if this is possible or not.
So either what you were implying is true (in this case I shall surely never buy another drive from your company). Or it is not true (bummer), but then the drive is defective.
However -- in both cases, for me it's the same result: Dump that Torqx drive.
I appreciate your efforts however.
Well then I am sorry you feel this way and I hope you enjoy your new drive.
Originally Posted by David.P
Sorry if I bring back this thread to life,
But I got the same exact problems using 128Gb Torqx2 under XP. Write performance is extremely low -much slower than a normal HDD- and system becomed almost unusable.
I was asking myself if someone from tech support completed the tests you was talking about with this SSD, since it seems obvious that this drive has a problem when used with XP. Note - it is 50% full,and was having no problem since used with 7. Unfortunately I have to use it with XP now - if possible - and it seem strange to see such a performance drop after only a few days of usage.
I will now try:
-TRIM BETA, just DLd from your website
-change cluster size from 4k to 64k and see what happens
Did you align the partition correctly? By default, XP partitions drives with incorrect alignment, which tends to screw most SSDs up.
Originally Posted by Falco75
No I didn't (forgot), but I don't think that this is the problem.
After using the TRIM BETA utility,the write speed peaks to 170-180 Mb/s,then progressively drops after some days of use to 30Mb/s and worse,making the PC unusable.
Tony Trim does not work at all and made speed worse than before.
Will try anyway to align the partition...any hints for the cluster size?
I doubt having a write speed of only 30meg per second would the PC unusable ... many hard drives from 7 years ago were that slow and they worked reasonably well with windows XP. Heck, my 100gig sandforce 1 drive only does 60-80meg per second writes and it is far from unusable.
That is, unless you have a heavy write load. If you do have a heavy write load, the Torqx 2 won't be able to keep up, and you will need a new drive. The Pyro/Pyro SE range handles heavy and persistent write loads better.
Seems like the align procedure worked - the drive looks much faster now. It was, however, a real trouble to align it without deleting data - and very risky for the data itself.
Managed to do it anyway - using a complex procedure found in the notebookreview.com forum,with GParted (and then CHKDSK from XP CD,after setting disk from AHCI to IDE, to repair the damage done to partition while aligning...)
The write speed at 20/30 mb/s was not the only trouble - there was also very much stuttering,in read and write.
Now using Crystaldiskmark I got the same values in read and write,but the SSD looks very fast - as it should be. Did not realize that disk align was so important - thanks for your help.