Raid0 and TRIM Support on Intel 11.5 OROM and Latest v11.5.0.1184 RST Drivers ?

[kensiko]

TRIM takes a few seconds to work, look at HDD led when deleting >4GB files.

[alan1476]

Exactly: You have to clean the registry from a faulty iaStorF entry, which has been forgotten to erase by the RST(e) drivers uninstall procedure.
You can easily verify, if you are effected by this issue, when you open the Device Manager and look into the driver details of your drives (right click > "Properties" > "Driver" > "Driver Details"). If you should find the iaStorF.sys there and within the Windows\system32\Drivers folder as well, you got the performance decrease by the SCSI filter driver iaStorF.sys, which is still active, but cannot really be used anymore.
Warning:
Don't try to delete the iaStorF.sys from the Drivers folder, because this will induce a BSOD at the next restart.
Solution:
Run the regedit command, open the HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Contro l\Class\{4D36E967-E325-11CE-BFC1-08002BE10318} or similar and delete the word "iaStorF".
After having restarted your PC everything should be fine again. You will even be able to delete the iaStorF.sys from the System32\Drivers folder.

Hi Fernando, I was wondering if you could give me some advice here, I currently own a Samsung 840 Pro 512gb, no raid, just 1 drive. I have x79 Asus Sabertooth motherboard. I tried the RSTe V3 drivers and they write capped my drive in a horrid way, less than 3mbps on the 4K. I uninstalled the V3e drivers and used the msachi driver and it was a dramatic improvement but not what I got on the first run of Anvil Benchmark. I looked in the registry and there was an entry for the IASTOR.F driver exactly where you said it would be.



Uploaded with ImageShack.us

After deletion my Anvil with the msachi driver was better but not close to the first run after a clean install of Windows 7 64bit. Now what IRSTe driver should I use or should I just continue with the masachi driver?

[Fernando]

After deletion my Anvil with the msachi driver was better but not close to the first run after a clean install of Windows 7 64bit. Now what IRSTe driver should I use or should I just continue with the masachi driver?

As long as you are running your SATA drives in AHCI mode, I would stick with the generic MS AHCI driver.
As an alternative you may try the 64bit Intel RST driver v11.2.0.1006, which I have modified for being usable even with X79 Chipset systems. You can find it here: http://www.win-lite.de/wbb/board195-...-raid-treiber/. Remember: You have to force the installation by using the "Have Disk" button.

[alan1476]

As long as you are running your SATA drives in AHCI mode, I would stick with the generic MS AHCI driver.
As an alternative you may try the 64bit Intel RST driver v11.2.0.1006, which I have modified for being usable even with X79 Chipset systems. You can find it here: http://www.win-lite.de/wbb/board195-...-raid-treiber/. Remember: You have to force the installation by using the "Have Disk" button.

Thank you so much. Now I have to put the drivers on a disc I guess. LOL. You are very nice. Thank you.

[alan1476]

As long as you are running your SATA drives in AHCI mode, I would stick with the generic MS AHCI driver.
As an alternative you may try the 64bit Intel RST driver v11.2.0.1006, which I have modified for being usable even with X79 Chipset systems. You can find it here: http://www.win-lite.de/wbb/board195-...-raid-treiber/. Remember: You have to force the installation by using the "Have Disk" button.

Well I forced the drivers you gave me and the only thing it did was change the name of the drive in the Device Manager from Samsung 840 to just plain old " Disk" it however had no other consequences, it could be I didnt do it right. Still using MSACHI, and its write capping the drive, not to the extent where its not fast, but using Anvils Benchmark, it went from a 5300 to a 4300. All the 4K write score were way low. This happened after a clean install of Windows 7 64bit, it was great before I ran the Windows accessment and horrid after the accessment, very wierd.

[Fernando]

Well I forced the drivers you gave me and the only thing it did was change the name of the drive in the Device Manager from Samsung 840 to just plain old " Disk" it however had no other consequences, it could be I didnt do it right.

After having installed the Intel RST driver v11.2.0.1006, did you check the driver of the Intel SATA AHCI Controller? Only this way you are able to verify, that the driver update has been successful.

[alan1476]

After having installed the Intel RST driver v11.2.0.1006, did you check the driver of the Intel SATA AHCI Controller? Only this way you are able to verify, that the driver update has been successful.

Nope, but I am on it now, thanks again.

[Fernando]

Nope, but I am on it now, thanks again.

And what is your impression regarding this Intel RST driver compared with Microsoft's generic AHCI driver?

[alan1476]

And what is your impression regarding this Intel RST driver compared with Microsoft's generic AHCI driver?

All I can say is that you should be paid by Intel for fixing the stuff they should be fixing. Thankyou Fernando, you are a gerntleman.

[alan1476]

INTEL Chipset Device Software drivers 9.4.0.1005

* Product: Intel(R) Chipset Device Software
* Release: Production Version
* Version: 9.4.0.1005
* Target: Intel(R) Xeon(R) Processor E5 Product Family/Core i7, Intel(R) C200/C600/X79 series chipset


Download

[Fernando]

INTEL Chipset Device Software drivers 9.4.0.1005

These "Intel(R) Chipset Device Software" packages do not contain any driver, but just information (=text) files. Nobody is able to use them as AHCI or RAID driver.
So there are no new IRSTe drivers v9.4.0.1005 out.

Nevertheless there may be good news for users with an X79 chipset mainboard: A new Intel RSTe RAID ROM version 3.6.0.1023 is available! Users with an X79 chipset system can download this brandnew RAID ROM module from >here<.

[felix_w]

Hi all,

In addition to Fernando's HxD method of verifying TRIM functionality, i would like to add this program i just found :

TRIMCheck

I'm just adding the post of user "Bad Machine" of theSSDreview forums, where i found it :

TRIMcheck: A FREEWARE TRIM TESTER
The title says it all. fsutil will tell you if TRIM is enabled or not, but it won't tell you if it is really functional. TRIMcheck will actually check if TRIM works or not.

This program has no GUI, it runs in a command prompt window. You must copy it on the SSD that you are testing and run it twice, waiting for at least 30 seconds to a minute between runs in order to ensure accurate results.

When executed for the first time it will write on specific sectors on the SSD. It will then delete the written data and exit. When you run it a second time a minute later, it will try to read the exact same sectors. When TRIM works you'll get zeroes on the sectors it wrote, plus a notification that TRIM works. If the data is still there you'll be notified that TRIM is not working. This is good for cases where certain software use cheap workarounds that cripple TRIM in order to ensure SSD compatibility. In such cases fsutil will still show that TRIM is enabled, regardless if it really works or not. With this proggie you'll know for sure.

Hope you like it, i just tested it with success on system (non-raid).


EDIT :

Since there has been some misunderstanding, i did not develop the tool, i clearly stated that i found it and gave a link to where i found it.

Developer is Vladimir Panteleev and source of the tool is CyberShadow's blog.

Both are stated in the program running window, first 2 lines.

[B Gates]

Hi all,

In addition to Fernando's HxD method of verifying TRIM functionality, i would like to add this program i just found :

TRIMCheck

I'm just adding the post of user "Bad Machine" of theSSDreview forums, where i found it :




Hope you like it, i just tested it with success on system (non-raid).

Pretty cool thanks felix

[Fernando]

@ felix:
Thank you very much for the informations regarding the TrimCheck tool, which has been offered by the user "Bad Machine" at "The SSD Review Forum."
I have just tested it with my Intel Z77 RAID0 array and this is what I got:

Just to verify the results, I executed the "Easy Trim Test" method by using a hex editor, which I had layed down >here<. Result: The sector, where the jpg file had been, zeroed out within a few seconds after the delition of the file.
So I am rather unsure, if "Bad Machine"s TrimCheck tool gives realistic results, if the SSD is part of a RAID array.

[kensiko]

Great program! I tried to test TRIM on any of my SSD using the hex method and I never succeeded, but now I got that:

Re-checking raw volume data...
Opening \\.\C:...
Seeking to position 39428247552...
Reading 16384 bytes...
First 16 bytes: 6F D2 64 F4 28 00 07 95 21 E3 15 2B D3 D7 61 C2...

Test file created and deleted, and continuation data saved.
Do what needs to be done to activate the SSD's TRIM functionality,
and run this program again.
Usually, you just need to wait a bit (around 15 seconds).
Sometimes, a reboot is necessary.

Press Enter to exit...


C:\>trimcheck-0.3.exe
TRIM check v0.3 - Written by Vladimir Panteleev
https://github.com/CyberShadow/trimcheck

Loading continuation data from C:\trimcheck-cont.json...
Drive path : \\.\C:
Offset : 39428247552
Random data : 6F D2 64 F4 28 00 07 95 21 E3 15 2B D3 D7 61 C2...

Reading raw volume data...
Opening \\.\C:...
Seeking to position 39428247552...
Reading 16384 bytes...
First 16 bytes: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00...
Data is empty (filled with 0x00 bytes).

CONCLUSION: TRIM appears to be WORKING!

Edit:
Negative result for my RAID array.

[felix_w]

As far as i can see, the tool uses a similar method with Fernando's Trim Test method, although it does it automatically.

Ιt may be easier , yet not known if 100% valid, for some non-experienced users, maybe not existing in XS, but out there, there are some.

[Ourasi]

@ felix:
Thank you very much for the informations regarding the TrimCheck tool, which has been offered by the user "Bad Machine" at "The SSD Review Forum."
I have just tested it with my Intel Z77 RAID0 array and this is what I got:

Just to verify the results, I executed the "Easy Trim Test" method by using a hex editor, which I had layed down >here<. Result: The sector, where the jpg file had been, zeroed out within a few seconds after the delition of the file.
So I am rather unsure, if "Bad Machine"s TrimCheck tool gives realistic results, if the SSD is part of a RAID array.

Hi, Ferando!

You have to run the tool to create testfile, run it again to trigger deletion, then reboot and run it again with the same *.json created before the reboot (reboot forces TRIM success every time).. It seems only singledrive TRIM's instant with this tool, the tool possibly is forced into a queue by the OS/drivers, since it's not a user-triggered emptying of the recyclebin ..

Intel 520 Raid0:


Crucial C300 Raid0:


Single 840 Pro OS drive:


Done on Asus P67 Sabertooth, all Raid0 volumes successfully TRIMMED..

*Edit: Updated with correct method...

[Fernando]

You have to run the tool, wait a few minutes or so, then run it again.

This tool doesn't work for me.
It doesn't matter how long I wait before I rerun the tool, even after a reboot: I always get the message "TRIM appears to be NOT WORKING".
But when I do the hex editor test, the related section zeroes out within 2-3 seconds.

[B Gates]

This tool doesn't work for me.
It doesn't matter how long I wait before I rerun the tool, even after a reboot: I always get the message "TRIM appears to be NOT WORKING".
But when I do the hex editor test, the related section zeroes out within 2-3 seconds.

Doesn't work for me either and I know TRIM is functioning

[Ourasi]

This tool doesn't work for me.
It doesn't matter how long I wait before I rerun the tool, even after a reboot: I always get the message "TRIM appears to be NOT WORKING".
But when I do the hex editor test, the related section zeroes out within 2-3 seconds.

Here is what workes for me every time: Put the tool on the volume you want to test: Run the tool, it creates the testfile and *.json, run it again imidiately to trigger deletion, do not delete the *.json. Then reboot, and run it again.. Works every time on my end on Win7...

First run creates testfile+json only, if you reboot now, it wont work since no deletion has been triggered. Reboot only after you run the tool the second time, as the second run triggers deletion. The third run after the reboot should see success as the sectors are zeroed out...

Anyway, I'm sure TRIM works for both of you, and I suspect there are some issues with this tool, since I to do not need a reboot or a long wait with the hex editor test, with hex editor it TRIMS instantly..

[Fernando]

Here is what workes for me every time: Put the tool on the volume you want to test: Run the tool, it creates the testfile and *.json, run it again imidiately to trigger deletion, do not delete the *.json. Then reboot, and run it again.. Works every time on my end on Win7...

Exactly this procedure doesn't work on my system (Win8 x64). Surprisingly the tool detects TRIM as working, if I wait minimum 30 minutes and then rerun it, but this is not a correct TRIM verification. It could be the SSD in-box Garbage Collection as well, which cleaned the SSD area within that long time.

[alan1476]

This tool doesn't work for me.
It doesn't matter how long I wait before I rerun the tool, even after a reboot: I always get the message "TRIM appears to be NOT WORKING".
But when I do the hex editor test, the related section zeroes out within 2-3 seconds.

It doesn't work for me either, but I do not believe this tools works at all. It might and Imight be doing it wrong but it doesnt detect TRIM on my machine and faustil says its running. Its says " indeterminate" wait 15 sec and reboot, well I did that and it keeps saying it. It wont work.

[Fernando]

Meanwhile I have done some intensive tests with the version 0.3 of the TrimCheck tool.
Here are the details of my test system:

1. Hardware: Z77 chipset MB with 2x256 GB Samsung 840 Pro as RAID0
2. OS: Windows 8 Pro x64
3. Intel RAID ROM: RST(e) v11.6.0.1702
4. RAID drivers: Intel RST(e) v11.7.0.1013

TRIM definitively is active within my RAIDed SSDs. That has been proven very often by having done the "Easy TRIM test method", which I have described >here<. After having completely deleted the JPG file the related section always zeroed out within 2-3 seconds.

Having these previous TRIM test results in my mind, I was very astonished, that I always got this message, when I ran the "TrimCheck" tool:


Yesterday I have done the following experiments:

1. I executed the TRIMCheck tool as Administrator and reran it 1, 2, 3, 4, 5, 10, 15, 30 and 60 minutes later.
   Results:
   Each time I got the following message:
   
   So the TrimTest tool didn't detect any TRIM activity.
2. I restarted the computer and let the TrimCheck tool check again.
   Result:
   
   So the reboot of the OS didn't change anything. TRIM has still not been detected by the tool.
3. Then I executed the Win8 "Optimizer" (former Defrag Tool), which sends a thunderstorm of TRIM commands to the related drive, and optimized drive C this way.
   A few seconds later I reran the TrimCheck tool.
   Result:
   Surprise, surprise - now I got this message:
   

These resuts verify, that the current version 0.3 of the TrimCheck tool is not able to detect any TRIM activity unless the release of TRIM commands by the OS has been pushed by a separate action.
Conclusion:
Contrary to the manually done Hex Editor method (see above) the TrimCheck tool is not able to detect "normal", that means unpushed TRIM activity.

My comment:
When I compare the action of CyberShadow's (Vladimir Panteleev's) TrimCheck tool with the manual Hex Editor TRIM test, I suspect, that the test file, which is copied by v0.3 of the TrimCheck tool onto drive C and deleted later on, is too small sized (16,4 KB).
Obviously the insertion and delition of a much bigger sized file (> 1 MB?) is needed to let the OS send out TRIM commands by itself.

[kensiko]

FYI I ran the tool again on my RAID array and now it is all 00 00 00. It's possibly due to the garbage collector, not sure.

[Fernando]

FYI I ran the tool again on my RAID array and now it is all 00 00 00. It's possibly due to the garbage collector, not sure.

It sometimes works for me as well.
If it shows "TRIM appears to be WORKING" within 1 minute after having executed the EXE file the first time, you can be sure, that TRIM is active. No Garbage Collection is able to clean SSD cells 100% within one minute.