Question O&O Software - Defrag 23 - purports to defrag (internally?) SSDs, such that they use a minimum of memory cells to hold your data. How does this work?

[VirtualLarry]

O&O Defrag 30: Maximum Performance for Your PC

O&O Defrag reorganizes scattered file fragments to speed up Windows, save time, reduce drive wear, extend disk life, and improve your chances of full data recovery.

www.oo-software.com

I thought that this sort of thing was "hidden" from the host system? How can O&O defrag, "internally defrag" SSDs? I would like to know.

@Billy Tallis

Edit: Possibly, I'm misunderstanding their new marketing materials. Anyways, I just remembered, on NT-based systems, there's a specific defragmentation interface used by all defrag tools. If O&O uses that interface, then they can't be doing what they (appear?) to claim that they are doing, at least above and beyond the filesystem block to "physical" host LBAs. (With device LBAs still hidden by the FTL of the SSD.)

 

[ralfy]

There's a more detailed explanation here:

NEW! O&O Defrag 22 Pro - Optimize SSDs and HDDs With The New SOLID method! - O&O Software

SSD's also fragment files when storing them. Data is often distributed to more memory cells than would be necessary. By defragmenting an SSD, it is possible to reduce the number of cells used to store file parts from 10 memory cells down to just 2 memory cells. Read more...

blog.oo-software.com

 

[Billy Tallis]

There's a more detailed explanation here:

More detailed perhaps, but also definitely more wrong. I can imagine how a defrag-like tool could offer some marginal performance improvements, but they're wrong to claim that they are helping SSD lifespan.

 

[hojnikb]

I can't possibly see how this would actually work on SSDs. They are claiming to defrag files, so that in their theory it would take less blocks than necessary.

Unless you have direct access to SSDs FTL (which they most certainly don't) you can't actually affect something like this. SSD manages it's data on it's own and obviously performs things like garbage collection, wear leveling and the likes internally, hidden from the system.


To me, this looks like some marketing bullshit for what might essentially just be plain old manually trimming. Or a very poor understanding of inner working of flash devices. This might make some sense, if data was actually mapped 1:1 as it is on a logical level. Then, you might argue, that having fragmented data, unaligned with actual pages and blocks on an ssd would take more blocks, than necessary.
For example, a 4k page/128k block ssd with a fragmented 22K file. Assuming all of the 6 pages are "fragmented", that would take 768K work of actual flash space.
Thats my dumb explanation anyway

 

[TheELF]

Unless you have direct access to SSDs FTL (which they most certainly don't) you can't actually affect something like this. SSD manages it's data on it's own and obviously performs things like garbage collection, wear leveling and the likes internally, hidden from the system.

It doesn't matter who controls it,SSDs have allocation unit sizes just like any other drive and each fragment of a file will take up this allocation size amount of storage on the SSD.
Yes after the defrag the SSD firmware will kick in and start doing their own maintenance but this will take some time to begin and also smart firmware will only bother with files that are often written to.

More detailed perhaps, but also definitely more wrong. I can imagine how a defrag-like tool could offer some marginal performance improvements, but they're wrong to claim that they are helping SSD lifespan.

Well if you use less cells then a lower percentage of cells will have wear so you get a higher life span but I doubt it's any reasonable amount.

 

[VirtualLarry]

Unless you have direct access to SSDs FTL (which they most certainly don't) you can't actually affect something like this.

I think that's basically the bottom line with SATA SSDs, and why I was so incredulous at their marketing claims.