Background optimization while HDD is idle

[quanthero]

I found a post on OCZ Forum which shows that there is an improvement after defragmenting an SSD, albeit a small one. Here's link in case anyone's interested:
http://www.ocztechnologyforum.com/forum/showpost.php?p=390702&viewpost=1.

However, these results are obtained on non-trim compatible SSD. There is absolutely zero performance improvement from defragmenting TRIM-compatible SSD on Windows 7:
http://www.anandtech.com/storage/showdoc.aspx?i=3667&p=2.

[cg]

Most of the newer SSDs today have lifespans that will vastly outlive the computer they reside in.  I read an article recently where they did some testing and based on the results said that even if you were to fully fill and wipe the SSD drive 5 times a day that the mean lifespan would still be close to 10 years.  I guess its because there are no moving parts and no heat generation.

Older Flash disks couldn't do this...

[jeroen]

There is absolutely zero performance improvement from defragmenting TRIM-compatible SSD on Windows 7:

I can't find that in that article. And if it says that, then the article is wrong.

p.s. Do not confuse memory block fragmentation (which is what TRIM deals with) with filesystem fragmentation (which is what MyDefrag deals with).

[quanthero]

There is absolutely zero performance improvement from defragmenting TRIM-compatible SSD on Windows 7:

I can't find that in that article. And if it says that, then the article is wrong.

It is obvious from table. After deleting all files on SSD (which forces TRIM), write performance is restored from 13.8 MB/s to its original level. So, since TRIM restores SSD performance to its original level, there is nothing to gain from defrag.

p.s. Do not confuse memory block fragmentation (which is what TRIM deals with) with filesystem fragmentation (which is what MyDefrag deals with).

I believe I don't. It just happens that defragmentation adresses SSD block issue indirectly (by consolidating free space) which helps write performance. But now that we have TRIM (which is a proper way of addressing SSD block issue), there is no need to address SSD block issue by any other indirect method (e.g. defragmentation).


Regarding SSD read performance, I have not seen any evidence that defragmentation at file system level improves read performance in any significant way.

[jeroen]

So, since TRIM restores SSD performance to its original level, there is nothing to gain from defrag.

Yeah, that's what I thought. You are confusing memory block fragmentation and file system fragmentation. TRIM restores the random read and write performance of the SSD. But it does nothing for performance loss because of filesystem fragmentation. Windows has to do more work when a file is fragmented, gathering all the fragments. There is significant overhead inside Windows, nothing to do with the hardware, and it is all the more noticeable because SSD is so fast.

Regarding SSD read performance, I have not seen any evidence that defragmentation at file system level improves read performance in any significant way.

Do you have an SSD disk? Several people here on the forum have reported significant improvements. I am not talking theory here, but actual hard results.

[quanthero]

Thanks for explanation, Jeroen! I indeed confused SSD block fragmentation and file system fragmentation. But after your explanation, the distinction is clear.