SSD optimizer finishes in 2 seconds

[idata]

Hi.

I'm using a 160 Gb G2 ssd with 2HD firmware on Windows 7 64 bit, AHCI. I had some problems with SATA drivers (lenovo update utility installed matrix storage 7.x which doesn't support TRIM) so I installed version 9.6 drivers. But I wanted to run the SSD optimizer (v1.3) to clean up after the old driver, but as soon as it starts it exits after 2 seconds. The result is passed and green, but i'm sure it didn't do anything. It has more than 100 Gb free space and according to SMART, 180 GB was written to it. I tried to run it with the old driver but it's the same.

Thanks

A.

[idata]

Sorry, but I got so many weird ideas about how manual trim is working and why is it taking 2 seconds now that was several hours before. I just want hear an offical answer, becuase I have a feeling it was over-optimized. Like if you have trim compatible drivers, than you don't need manual trimming. But this is unture. One more example: what if I have a dual boot system with xp? When will the data deleted by xp be trimmed?

If trim is so quick (100 GB in 2 seconds?) why did it take hours before? If it is done in the background by the firmware how is it able to create the free/occupied bitmap in 2 seconds from MFT data with hundreds of thounds of files and folders. (if you try an analyse with any defrag tool, which is doing the same, you will see it takes much more time. For me it's about 15-20 seconds)

What happens if I restart to XP while this "background trim" is working? Etc-etc.

If Intel made this unbelievable optimization like doing the same job in 2 seconds which was hours before, why it was not advertised, not even mentioned anywhere on the internet?

Anyway, I asked the offical support about this, if I get a technical answer I will post it.

[mmokk]

acider wrote:

The problem is not what was deleted by the old OS. When I installed win7 and repartitioned the SSD everything was "deleted". But there was certainly no trim for this "delete".

when the partition is formatted, it is trimmed.

the exception is if the ssd is part of a raid array, or if it is connected to the sata controller in raid mode (non-member raid) and you use the default windows 7 raid drivers instead of loading the intel rst 9.6 drivers, since the raid drivers included with windows 7 does not support trim passthrough.

[mmokk]

acider wrote:

Sorry, but I got so many weird ideas about how manual trim is working and why is it taking 2 seconds now that was several hours before. I just want hear an offical answer, becuase I have a feeling it was over-optimized. Like if you have trim compatible drivers, than you don't need manual trimming. But this is unture. One more example: what if I have a dual boot system with xp? When will the data deleted by xp be trimmed?

if you boot into any os that does not support trim and delete files from the ssd, then it will not be trimmed. when you then boot into windows 7 (or any other os that supports trim), it will not be aware of what the other os has done. the untrimmed data will stay untrimmed.

if you do this often, then you may find it beneficial to run trim manually.

[idata]

"if you do this often, then you may find it beneficial to run trim manually."

That's why I opened this thread. I can't run it. Or I just don't believe it's able to do the manual trim in 2 seconds. I think it's not doing anything. If I'll have time I will proove it by booting an XP, writing a big file on SSD, deleting it, booting back to Win 7, running SSD optimizer, then searching the file's content with a disk editor (winhex). I'm sure it will be there...

[mmokk]

acider wrote:

If I'll have time I will proove it by booting an XP, writing a big file on SSD, deleting it, booting back to Win 7, running SSD optimizer, then searching the file's content with a disk editor (winhex). I'm sure it will be there...a better way of testing it is to boot into xp, write a big file to the ssd, delete it, and run a sequential write benchmark. the write performance should be significantly lower than advertised specs. then run the ssd optimizer (either under xp or win7, whichever works for you) and run the write benchmark again. write performance should be very close to advertised specs.