I might be wrong on certain points.
. The asstahci drivers appeared with v3.1.6.0
. The readme.txt file contains good info on the subject
. The setup program is in charge of the procedure to determine the suitability of driver/os combination
. The directory “Driver_Win10” only contains “stahci” directory (ie asstahci32.sys / asstahci64.sys)
. What defines the OS support is the CAT file (ie if you look into the “Driver” directory there is indeed a “stahci” directory which only differs from the aforementioned one by the CAT files)
. So yes it might be possible to install stahci on others OS than Windows 10, but this is not the default behaviour of the installer.
. DeviceID are the same (why not ?). Not everyone plays like Intel :->.
. INFs are nearly identical because it’s not an architectural change they are both Storport drivers (ie not SCSI Miniport drivers like in XP)
Hi there,
has anyone else experienced the dreadful drive spinup, spindown, then spinup again-problem during Windows shutdown using the newer ASM1061-drivers (version 3.x.x.x) ?
I’m on Win7 64bit AHCI (using Port Multipliers quite extensively), and I didn’t see that behaviour using the older drivers (v2.x.x.x).
Is there a known fix for this, like some registry-setting or so ?
Someone else reported this problem here before, but using Intel RST drivers.
I don’t have any other BIOS settings that I could try, I use ASM1061 on PCIe-cards without any settings available.
Any input appreciated !
shroeder,
Interesting, you state that Win 7 uses asstor32.sys/asstor64.sys drivers. Using Device Manager to install/update the drivers the only options I have for a driver is the asstachi64.sys driver. My system is X64 so the 64 bit version makes sense to me. I am not running a raid setup which I was under the understanding that the asstor.sys drivers were for. Am I missing or not understanding something here?
My write-caching parameters were identical between the benchmark runs posted.
I am aware that various revisions will yield different results and that is at the very root of my question. With Fernando’s explanation that the only difference in the v3.1.8.0 drivers is that of digital signature do you have any idea why my system BSOD’d when installing this version? Is that or could it be the result of installing the ASSTACHI driver rather than the ASSTOR driver? If my system truly should be running the ASSTOR variant how do I get Device Manager to list it as a valid driver for my system?
EDIT: I do stand corrected Sir! I retraced my steps and found that I had not copied the correct driver folder for my PC so that I had access to the ASSTOR drivers! DUH! do I feel dumb! I have 30+ years experience with this stuff and you would think one would not make such a blunder! Oh well, got things sorted now and correct drivers installed. Will test when time permits. A BIG thanks for the wake up call!
shroeder, Fernando,
Thank you guys a bunch! Your explanations give me my confidence back. You see, I have 6 systems I run, 3 of them are Win 7 and 3 are Win 10. One of these systems is a dedicated Test Bench machine in which I have an Icy Dock 4 bay 2.5 inch X 5.25 adapter. This allows me to swap in and out OS installs to this machine. In moving to Win 10 Pro I noticed that data transfer performance between 2 SSD’s was very poor whereas transfer from an SSD to an HDD was much better. This led me to start looking for better drivers for the controllers for my board (Asus Z97 Deluxe) which sports an Intel and Asmedia SATA controllers with the Asmedia having SATA and SATA Express available.
The benchmark numbers I get from the Asmedia controller are a bit better than that of the Intel controller but not enough to write home about. The issue I am seeing with the SSD to SSD transfer in Win 10 is that the average transfer rate for a large amount (20GB) of random data runs about 55MBps. When the same data is run from SSD to HDD I am seeing averages of 120MBps+ which is expected (HDD is Seagate 1TB Barracuda).
The SSD source drive for the data is attached to the Intel controller and the destination SSD drive is attached to the Asmedia controller. I have changed these from one controller to another in every conceivable configuration and nothing makes any difference. So I am now wishing to test if Win 7 shows the same behavior as Win 10 does. Am trying to find out if this is an OS issue, hardware issue, or what!
At least with your explanations about the driver differences above I can see that my mistake was an honest one as not realizing that these different OS’s used different driver variants caused me issue. Am very glad I registered here as this certainly is most helpful!
Thanks again!
@ pupvogel (what a nick name!):
Welcome at Win-RAID Forum!
AFAIK the only possibility to get rid of that issue is to change the driver, but I am not sure about that.
Regards
Dieter (alias Fernando)
Update of the start post
Changelog:
-
ASMedia 106x SATA3 AHCI/RAID drivers:
- new: 32/64bit ASMedia 106x AHCI/RAID drivers v3.1.8.0 WHQL for Vista and Win7
Thanks to our Forum member shroeder for the link and for his statement regarding the OS support.
Any feedback is much appreciated!
Good luck with these ASMedia AHCI/RAID drivers!
Dieter (alias Fernando)
Hello again,
Thought I would post an update of my experience with the Asmedia 106x drivers 3.1.8.0.
Using this latest revision I see a dramatic increase in write performance running this driver under Win 7 X64 and Win 10 Pro X64. To recap I was experiencing poor write performance between 2 SSD’s using previous revision drivers whereas this was not the case when writes occurred between an SSD and an HDD. Although writes between 2 SSD’s is still not where I think they should be I did see performance increase by almost 60% from an average of 55MBps to 85+MBps. These are not big numbers and I am not expecting big numbers as these drives are small capacity entry level SSD’s. Even so, performance should be better.
Benchmarking showed nice gains on both read and write performance with the new drivers but real world results even though a gain of 60% was achieved, is just not where it should be. In contrast write performance from either SSD to a 1TB Seagate Barracuda drive averages 164MBps. These are real world results using file copy and not benchmark results.
I have a Samsung 850 EVO 250GB drive I am going to add in the mix and see if results are better. I am leaning towards this issue being that of drive controller design issues with these cheap entry level SSD’s.
UPDATE
Just to update my last post I did add 2 other SSD’s to the mix of drives and can confirm that drive/controller design is at the root of my performance issues. You get what you pay for!
Update of the start post
Changelog:
-
Marvell 91xx/92xx SATA3 AHCI/RAID drivers:
- new: 32/64bit Marvell 91xx/92xx SATA3 AHCI/RAID Miniport Drivers v1.2.0.1038 for Windows XP and Windows Server 2003
Thanks to our Forum member shroeder for the link and for his informations regarding the OS support of the different Marvell SATA3 drivers.
Any feedback is much appreciated!
Good luck with these Marvell AHCI/RAID drivers!
Dieter (alias Fernando)
hELLO Fernado Sir ,Can you please make Intel NVMe drivers v1.3.0.1007 available for XP 32bit by modding anway
Please please
Thanks for Raid drivers
Does XP on A Z170/97 possible with AHCI support.
That would be a very interesting task, but I don’t think, that I will ever offer such modded NVMe driver.
Reasons:
- It is rather unsure, whether an old OS like Windows XP is compatible at all with the absolute new NVMe protocol.
- Windows XP needs a textmode driverpack to be able to boot off it. Since the required file named txtsetup.oem doesn’t exist for any NVMe driver, there is no source file available for any modification. It has to be constructed from scratch.
- Contrary to the MS standard NVMe driver the Intel NVMe drivers have an additional SCSI filter driver named IaNVMeF.sys. I doubt, that the loading and the usage of this SCSI driver will work with Windows XP.
Yes, if the system drive is connected to any Intel SATA port and you load or integrate the "Universal 32bit Intel RST AHCI+RAID driver v11.2.0.1006 mod+signed by Fernando".
Hello
If I install the driver pure 32/64bit Samsung NVMe Windows Driver v1.4.7.6 WHQL (with the .inf) I have a problem with detection USB Drive (Windows 7 and 10)
No problem with the .exe (with the SM 951 NVme)
Same for you ?
Thanks
@ tistou77:
Thanks for your report.
What happens, when you install the pure 32/64bit Samsung NVMe Driver v1.4.7.16 WHQL instead of the older v1.4.7.6?
If the newer version shouldn’t show your reported issue, I will remove the old driver from the start post.
Hi Fernando
I have not tested the 1.4.7.6, 1.4.7.16 only (with X99 plateform)
If I install the driver for the device manager (.inf) when I plug in a USB key, it is not recognized and is "freezing" the PC (the PC does not turn off, etc …)
as soon as I remove the USB key, everything is normal
Thanks for the clarification.
Meanwhile I have added your find to the start post.
You’re welcome 
Someone with a X99 platform install the driver by the device manager (.inf) and everything works well ?
Which AHCI driver is most suitable for use in Win 10 x64?
Is it, or is it not recommended to use the Win8 driver?
Can’t really seem to find it!
There is no difference between Win8/8.1 and Win10 regarding the recommended AHCI driver.
I don’t know any Win8 compatible AHCI driver, which is not usable with Win10.
Thanks Fernando!
Well i gave it a try,and found that the performance gain is zilch… just under 1%
Which could as well have been a fluctuation between benchmarks!