Corsair Force Series 3 60 GB 130 MB/s

Full story short: just bought one Corsair Force Series 3 60 GB, chipset is X58 if u ask. By the way it's minimum transfert rate ( READ ) out of the box was exactly 130 MB/s. A little disappointed, I've downloaded the firmware update 1.3.3, updated the thing and made an usb stick for erase the SSD. Got back 235MB/s.

I tought well, nice.. I don't need to go out and replace the drive. Moved back few game directory, Crysis, Crysis 2, Battlefield 3 and Metro 2033 and this is what I've found:

 

http://img830.imageshack.us/img830/4950/ssdmp.jpg

 

( This image rapresent the drive, after copying 44,1GB of data )

 

Serial is 11456511000010680835.

C state is disabled along C3/C6.

The drive have right now 11,6 GB of 00s left ( aka EMPTY ).

The drive is attached on the ICHR.

 

Is this for real?

Test with ATTO.

I tought that one image could explain ten thousend word..

ATTO doesn't exactly provide a real output in term of actual performance.

I'm asking if these drive have an awful READ speed once a bit is written to it.

Is this expected? is it normal for this SSD to READ actual 64 61 74 61 @ 130MB/s?

 

 

As requested ATTO:

 

http://img14.imageshack.us/img14/4381/attoq.jpg

 

Which is normal since it's running on 300MB port, alas.. I will and need to use this drive as drive EG: STORING DATA not BENCHMARKING 00s.

 

:wtfeh:

ATTO is what the industry uses so that is what we use. There are simply too many other flaky half baked benchmarks out there that are not made for testing SSDs.

 

A benchmark is just that, a benchmark. If the drive is running well according to ATTO, and it's stable, just use it and enjoy the speed.

Thank Yellowbeard I do understand, anyhow mine question remain. Can you bear with me for a little longer, please?

why this SSD once empty is actually working under it's spec ( 250 MB/S ) and when some data is present the READ speed is degraded to 130 MB/s? ( see image first post )

 

http://img27.imageshack.us/img27/4527/emptyssd.jpg

 

( This image rapresent the drive, after a secure erase. Drive is EMPTY )

I am sorry but no, I cannot comment on how/why this particular benchmark shows this result. If you have questions about the results, I'd suggest you consult the makers of the benchmark.

HDTune uses a sort of LBA query.. or maybe better to say "pinging".. to test bandwidth.

 

Now because these particular controllers use internal compression.. they will return lower reading speeds on the areas of the drive that contain data. The less compressible the data?.. the lower the result. Some SSD's don't drop speeds in that benchmark as much but all will respond in similar fashion when comparing a drive with stored data to one that's empty. In essence, it will show you exactly where the data layout resides on the drive.

 

That's why HDTune not usually recommended for SSD's as the results often mislead people to thinking there's an issue when there is not.

 

PS. @Yellowbeard. I tried to reply to your PM several times(over a few days time) but the system says I can only respond once every 20 minutes. lol

I can name at least 4 SSD which doesn't drop like that once you copy a file on it.

Again, zero filled file does't really exist on mine idea of normal operation usage, sure.. I can open a CMD and type;

fsutil file createnew 3GB 3221225472

and then enjoing moving this usless "data" to and from this very new SSD.

Copying an incompressible file to and from it won't goes past 130 MB/s ( READ ) and 53 MB/s ( WRITE ).. you know, from the OS GUI.

So, for the very first time disappointed at last corsair.

 

Anyhow, it was a gift from mine wife, so imagine how she is seeing me here debating this.

Hope you get the fact stright and easy acessible, otherwise I will think twice next time before I had to purcase or raccomand a new PSU/Memory or the brand in general.

 

Kind regards