SSD underperforming - im new to ssd -s

[cheti]

So i bought a new corsair force gt 90gb ssd for windows and games - eveyrthing is working EXCELLENT and speed is insane.
On the box - (i dont remember correctly, but something about) 500mb/s reads and almost the same for writes.

Now the thing is - when i test i get about 200 reads and 100 writes.. whats wrong ?

What do i have to disable/enable make sure of ?
I am running in AHCI mode, im using a sata3 cable and its in the fourth sata port...
It has windows and games on it, its running at 34C when playing BF3.

Any suggestions ? thanks!

 

[cin19]

which program do you run on the benchmark?

 

[cheti]

AS SSD-Benchmark

 

[cin19]

1) do you upgrade the firmware from Corsair? http://www.corsair.com/us/support/faq/solid-state-drives/
2) how much free space is on your SSD?
3) do you follow the "useful SSD articles part2-Tweaks / Optimization? http://www.tomshardware.com/forum/270102-32-useful-articles-part

 

[cheti]

1) Yes, i did it when i had my previous windows on - do i have to do it again after a fresh windows 7 install ?
2) 20 GB out of 90GB.
3) Nope, reading it now

 

[geofelt]

1) On AS SSD the report will show the size of the ssd. If it is in green, that is good, you will be aligned on a 4k boundary. If it is in red, then you will be unaligned, and will not perform as well.

2) Is the 4th sata port a 6gb port? If not, switch things around.

3) 90gb is not that large. Benchmarks are done on drives which are largely unused. As you fill up the drive, there will be some performance impact. If the drive is >80% full, there might be some impact there.

4) The real value of a SSD is in the random I/O speed. A characteristic which you have already experienced.
It is probably not worth spending too much time worrying about synthetic benchmarks.

 

[geofelt]

re: tweaks...
Don't bother. Just my opinion.

 

[USHypertraxx]

LOL, i have the Corsair Force GT 120gig and same problem... The reply's however are funny as they not helping.

What chipset do you use? I have x56 or 58. Anyways the benchmarks are based on x78. And i read some more stuff on corsair. I'm very disapointed aswel. And i won't buy another SSD of corsair. This due to you need to plan your computer based on their ssd, Corsair said.

Also @cin19
You are the most common guy that you find on forums. 'did you update the firmware'?
READ THE CORSAIR WEBSITE FFS, it says DON'T update unless something really weird is going on, as there is a high change of losing data.
Really READ the website before you tell you genius stuff on forums. You probebly don't even own a Force GT....

 

[cheti]

Well, im happy that its a huge upgrade to my 5400rpm hdd and that my bf3 and windows loads in seconds, not minutes. But still - it bothers me that im not getting the full performance of it.

 

[USHypertraxx]

Well, im happy that its a huge upgrade to my 5400rpm hdd and that my bf3 and windows loads in seconds, not minutes. But still - it bothers me that im not getting the full performance of it.

True it does feel faster, even than a 7200rpm.
BUT
however,
If you own a WD Caviar BLACK...
You get about 2TB for the same price, and you won't feel the difference
As it can read up to 190mb/s... If only i knew the box speeds are almost impossible to reach...
Believe ive bin on the phone with Corsair for hours, demanding my money back.

 

[cin19]

i think you need tweaks your SSD, and i don't know how many programs in on your SSD. Also you can use the CCleaner or Advanced SystemCare Free 5 from cnet.com

Also @ USHypertraxx
i don't have Corsair Force GT but the reads and writes down from 500mb/s to 200 and 100 is NORMAL, do you read that first?

 

[Razec69]

You won't get the full performance.

With doing a lot of research those benchmarks of 500MB R and 450 Write are complete and utter BS.

Yes in a controlled benchmark or synthetic benchmark as many will say you will get some amazing speeds but in real world situations you won't get nowhere near the advertised speeds.

As you can tell y ou already see a huge imporvement over magnetic disk HD's. But as others are saying you won't get as big of an improvement as larger drives.

I myself always wanted to go with a 60GB SSD for my OS, but then I said I wanted to add some programs and games and then opted for a 128GB SSD.

Then with research found out that anything less than 256 is not as good. So with that I am now going to find a good 256GB SSD.

It sucks but that how it is. Read around the forums and you will get a lot of insight on how SSDs operate.

 

[cheti]

Alright thank you!

 

[USHypertraxx]

i think you need tweaks your SSD, and i don't know how many programs in on your SSD. Also you can use the CCleaner or Advanced SystemCare Free 5 from cnet.com

Also @ USHypertraxx
i don't have Corsair Force GT but the reads and writes down from 500mb/s to 200 and 100 is NORMAL, do you read that first?

Ye thats why i asked for you chipset version. As you have 200/100 and mine goes up to 350mb/s read and 400mb/s, most common speed is 300/300mb/s for me... Still not even close to 555mb/s as my GT box says..

 

[cin19]

OP has 990fxa-ud3 mobi that support SATAIII (6GBs) and the speed down from 500mb/s before to now 200mb/s.

i think the problem is there are too much staff (about 70Gb) in the SSD.

 

[MEMOFLEX]

I know i struggled to hit advertised speeds on a sata3 SSD but that was due to the Marvell controller (9128) on my board not being the best regardless of space available. Sent it back and stuck with my Vertex 2e on sata2 which is still more than quick enough. What speeds you get on sata2?

 

[RetiredChief]

WOW
1) SSD manufs LOUDLY proclain Sequencial performance based on ATTO (a bench mark designed for HDDs).
.. A) Sequencial performance is the least important parameter for an OS + Program SSD, its the small file random performance that are important.
.. B) ATTO used data that is readily compressable, Sandforce controller based SSDs LOVE this, even though it it is NOT realistic. AS SSD (designed for SSDs) uses data that is already compressed - Hense the difference. Real life is closer to being Uncompressable on a OS + Program drive, So AS SSD is a Better benchmark even though you do not get the "advertised Sequencial performance". As Stated, Sequencial is of MINOR importance anyway.

And Yes chipsets make a difference also. The Marvel based MB controllers do NOT perform as well as the Intel controller. Intel also updates and Improves their driver more often than Marvel, or for that matter AMD. There is a reason Most reviews of SSDs are done on an Intel based Setup.

Added marvel = most 3rd party controllers.

 

[USHypertraxx]

WOW
1) SSD manufs LOUDLY proclain Sequencial performance based on ATTO (a bench mark designed for HDDs).
.. A) Sequencial performance is the least important parameter for an OS + Program SSD, its the small file random performance that are important.
.. B) ATTO used data that is readily compressable, Sandforce controller based SSDs LOVE this, even though it it is NOT realistic. AS SSD (designed for SSDs) uses data that is already compressed - Hense the difference. Real life is closer to being Uncompressable on a OS + Program drive, So AS SSD is a Better benchmark even though you do not get the "advertised Sequencial performance". As Stated, Sequencial is of MINOR importance anyway.

And Yes chipsets make a difference also. The Marvel based MB controllers do NOT perform as well as the Intel controller. Intel also updates and Improves their driver more often than Marvel, or for that matter AMD. There is a reason Most reviews of SSDs are done on an Intel based Setup.

Added marvel = most 3rd party controllers.

True,

I will NEVER EVER buy an Marvel based MB again. ITS CRAP.