One thing I do find curious in the advice given is the following point: 

 5. Maintain At Least 15% of Free Drive Space.

I thought this was standard advice for old fashioned hard drives, because when they got very full (and unless defragmented) the drive heads would physically have to work very hard shuffling back and forth to get the various bits of data.

 

By contrast, this is precisely the sort of thing SSD's are not meant to be bothered with (fragmentation), and I looked rather forward to using as much of the capacity of the drive as I chose to without having to keep an eye on leaving a margin.

 

What is it that causes SSD's to slow too as they approach full capacity, or is this just a function of Windows, or something else?

Regarding the drive not being configured in Ultra Direct Memory Access (UDMA) Mode 5.

 

A previous thread suggested toggling to "PIO Only" mode, rebooting then reselecting "DMA if Available" and rebooting again.  I'm not sure why this new help file does not mention this, as it is the obvious source of a slow drive if it's stuck in PIO mode.

 

I tried the toggle, but it didn't work for me.

 

Solution to if your drive is stuck in PIO mode

 

I found the trick was to force Windows to redetect the drive:

 

 

---

This is mainly for XP users but is informative to all.
If you are sure your drive is DMA but XP seems stuck on PIO. First check in device manager under IDE controllers that you have it set to "DMA if available" for your drive. If you reboot and recheck and it is still stuck on PIO which mine were then go to run and type in regedit. Find the following key:

HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Control\Class\{4D36E96A-E325-11CE-BFC1-08002BE10318}\000X

000X "X" being the number of the controller such as 0001 for primary and 0002 for secondary. Under this location find the key "MasterIdDataCheckSum" or "SlaveIdDataCheckSum" depending on which one your drive is and delete the value for the key. this will force XP to redetect your drive after reboot. Please be careful when editing your registry.

---

 

Sorry if this is the wrong place to post it, but I would hate for somebody to have to spend the best part of 24hrs like I did trying to resolve the problem when the answer could be right here. :)

Well, the physical process of heads flitting across the disks trying to piece together fragmented pieces of a file causes delays, but there is a logical process that governs this physical process too.  This logical process is the process of putting together the fragments of the file based on the MFT entries for that file.  As fragmentation increases, this logical process can result in delays too, even though the physical process of collecting the fragments once they are logically strung together may be much faster.

 

However, leaving at least 15% empty is a red herring in the case of both kinds of drives.  The important thing is to ensure defragmented drives.  And if the windows defragmenter will not defragment your drive just because it does not have at least 15% free, ditch it and try something like the freeware Auslogics Disk Defragmenter.

I was under the belief that the SSD's low level controller scatters data all over the place anyway as part of it's "wear leveller" regieme, so defragmentation of an SSD is sort of a red herring regardless of what the OS tells you about the state of fragmentation.  I'm personally not going to bother with defragging - I never have for my conventional hard drives and havn't regretted it, though that might be more a factor of the apps I run.

 

I think there is a lot of mystery with SSD's at present, I blame the manufacturers slightly for being a little secretive, but hopefully this situation won't last forever!

You are correct.  As a result of wear leveling on ssds, information is placed somewhat evenly throughout all nand modules.  In fact, without your knowledge, it is even moved around to ensure even wear leveling which makes defragmentation of a disk unnecessary.  It does the job without your knowledge.

 

On a traditional HD, information is placed at a certain sector and remains there.  Everytime it is accessed, fragmenting of that block may or may not occur.

 

Also regarding the 15% free space theory, I threw in my Memoright 32Gb SSD and decided to check on this making my remaining space around 5-6%.  I did not experience any slowdown whatsoever.  Keep in mind that my system is optimized and does not access the disk for things such as pagefile and system restore copies which may be what many are experiencing.

hi how are you doing? my laptop is b130. I want to upgrade my hard drive from 40GB to 120GB. I don't know if my laptop support to it. if yes what is max for my laptop hard drive. Thank you very much.

Ive recently got one of these SSDs and tried using it on my dell xps 1710. The problem im having is that it only seems to be using UDMA2 which is limiting the performance. Im not sure if its the drives fault as I only have one other system to test it in and that works fine at UDMA4 with this drive.

 

Running windows vista 32bit, updated to SP1, and also updated the chipset and sata controller drivers without any luck. The redetect registry key posted above must only apply to windows xp as i dont have a key matching "MasterIdDataCheckSum".

 

Any help would be very appreciated as I dont seem to be getting any responce from Samsung.

 

Thanks.

 

(Im also using BIOS version A06, havent updated to A07 as it doesnt seem to have any fixs relating to this.)

The registry key hack I posted was only one way to get to the same result.  Most other peoples suggestion was to delete/uninstall the controller in device manager (to be more precise, this is the hardware item which shows the UDMA2 value when you goto it's properties).

 

When Windows restarts, it will redetect the hardware and do the same thing as my registry trick.  However, I am uncertain if this will work with Vista, or indeed your particular problem as being stuck in UDMA2 instead of PIO suggests it could be something else.

 

There is a whole wealth of information out there for people with these kind of problems, but more often than not it's for DVD drives and that kind of thing (all of the google pages which helped me were about these).  I had a very quick look with Google and found this, which may or may not be of help:

 

http://forum.notebookreview.com/archive/index.php/t-104834.html

 

One poster says Vista could be a sticking point, another posts a registry trick... so would be interesting for the community if you post back with any results. :)

It's very possible that you Samsung issue is device related.  Seems to be a trend with them. 

 

 

However, I wouldn't think that the difference between ATA-66 (UDMA 4) vs. ATA-100 (UDMA 5) would be that significant anyways.  The bottleneck has always been the bus.  We didn't see, but about a 1 ~ 4% increase in performance with hard drives during that transition.  I would just like to see the write strategies improve.  I know that it doesn't write just a single pseudo block, rather it erases then writes which induces more lag time during writes.

 

We also didn't see that much of an increase between ATA-33 (UDMA 2) vs. ATA-66 (UDMA 4).

 

I would make sure you have the latest BIOS installed on your system if you have one of these drives. 

Ive been doing some looking around other forums and found that this problem is common with other dell laptops. At the moment dell and samsung seem to be passing the blame at each other but nothing is getting fixed as of yet. As for Ryanh UDMA2 is ATA-33 which is limiting me to  22-28MB/s in HD Tune and other benchmarks, where as most people are getting 54-62MB/s with the same drive.