I’ve just noted this in first post:
As I said, I can see one "PCI bus" named device with the same Hadware IDs as Switch - I can see it before my installation of USB drivers.
Could that mean that my "PCI Express Root Complex" could be somehow already "installed" on fresh install of Windows?
No, this is an ACPI device and not the Intel “USB Switch Device”.
After having updated the Intel USB Switch Device named “PCI Express Root Complex” you will get a second device named “PCI Bus”. To find out, which one is the Intel “USB Switch Device”, you should check the HardwareIDs of both listed devices named “PCI Bus” (VEN_8086 means Intel).
@100Pier
Thank you very much for your tests. It seems that windows 10 drivers for the xhci controller do the job for the usb2.0 devices.
This week I tried on another computer I have at work: MSI X99A SLI PLUS (LGA 2011-v3, Intel X99, ATX) motherboard. Same Windows7 x64.
If I plug a USB2.0 device in the front panel I can see that it appears under Intel(R) USB 3.0 eXtensible Host Controller, AND IT WORKS!
So, I assume that the Asrock x99 extreme motherboard have a real issues to manage correctly the xhci Intel controller.
This morning I just went to the shop, to bring back my motherboard…end of the story with this motherboard.
I am quite sure that they will not "repare" it.
Thanks guy for your help, I will tell you, what they decide at the end.
Jean
The thing is - I already have two "PCI Bus" devices. Here is my screenshot with Hardware IDs:
I’ve asked @100PIER , who also has X99 platform about his PCI Bus devices - he has one "PCI Bus" device and one "PCI Express Root Complex" device. Here is his screenshot wiht Hardware IDs:
His "PCI Express Root Complex" looks exactly like one of my "PCI Bus" devices. He has Windows 10, I have Windows 7.
@Fernando : I think it could not be the matter of X99 but it could depend on Windows version.
Maybe “PCI Express Root Complex” name does “not exist” on Windows 7, and Windows 7’s default name for this device is just “PCI Bus”?
From your point of view - if Hardware IDs are the same then devices should also be the same? In other words: can the same Hardware ID be used by different devices?
This may be possible. I cannot check it, because I only use Win10 on my machines.
Yes. When you connect 2 identical device models to the same Controller, both devices have the same HardwareIDs.
Has anybody tested the new ASMedia USB 3.0/3.1 drivers? If yes, don’t forget to post a short report.
I have it installed, no issues thus far, but no performance tests, don’t have anything that will really stress it (a few Intel 310’s in RAID 0 I guess), what kind of tests you looking for?
@Fernando
I have already posted #1431 some ASMEDIA USB3.x test results using the last driver version. I can do more in depth tests with different USB3.0 keys.
@100PIER :
Yes, I have seen your test results and hit the “Thanks” button, but I haven’t found there a benchmark comparison between the “old” ASMedia USB 3.0/3.1 drivers 1.16.41.3 and the latest v1.16.50.1 ones.
If the performance of the v1.16.41.3 drivers should still be noticeable better, I will leave the related chapter of the start post as it is now, but if there should not be a big difference, I will remove the old drivers.
Here are comparative tests done MS driver / ASMedia driver for 2 devices (USB2 and USB3).
MS driver is for reference bench.
Additional test done also with a USB3.1 device connected on USB3.1 ASMedia port:
The TRIM fonction does not work for the attached SSD on external Drive case USB 3.1 compliant (whatever the ASMedia diver version is used:
I observe slight better performance on ASMedia 3.1 ports than on Intel 3.0 ports. (the machine has none Intel USB 3.1 ports).
This "PCI Bus" device @ ACPI\PNP0A08 couldn’t get off my head so easily so I spent last two days browsing the internet, trying differents things and in the end I can say I think you need to do a small update.
During my searches I was lucky enough to find this link: https://superuser.com/questions/719079/u…el-usb-3-0-xhci
As it says "the Intel(R) USB 3.0 Host Controller Switch Driver is not installed" could cause "USB 3.0 ports function as USB 2.0 ports" so my first idea was to test the speed of my ports. The problem was I had not any USB 3.0 device, so today I bought SanDisk Ultra Flair 3.0 pendrive. I connected it to five different USB ports and tested write/read speed. I used Far Manager to copy a set of few 4GB size files (from HDD to pendrive and from pendrive to HDD) and I observed the results in Windows Resource Monitor.
Here are the results:
USB 2.0 Intel (front): W/R: 10 MB/s / 36 MB/s
USB 3.0 Intel (front): W/R: 10 MB/s / 36 MB/s
USB 3.0 Intel (back): W/R: 10 MB/s / 36 MB/s
USB 3.0 ASMedia (back): W/R: 10 MB/s / 36 MB/s
USB 3.1 ASMedia (back): W/R: 6.5 MB/s / 36 MB/s
As you can see all of them have 2.0 speeds, even ASMedia ones. I do not know why USB 3.1 had only 6.5 MB/s write.
So (still following the link) I decided to take a risk and install this third driver from Intel USB set (Switch Driver) to see if there will be any improvements.
I pointed "PCI Bus" device to iusb3hcs.inf - Windows detected "some suitable driver", installed it and prompted to reboot PC.
After restart, under Device Manager > Universal Serial Bus Controller two new devices appeared: "USB 2.0 MTT Hub" and "USB 3.0 Hub". By comparing the screenshot from before Switch Driver installation, I think one "Generic USB Hub" device disappeared.
Now tests:
USB 2.0 Intel (front): W/R: 10 MB/s / 42 MB/s
USB 3.0 Intel (front): W/R: 10* MB/s / 145 MB/s
USB 3.0 Intel (back): W/R: 10 MB/s / 125 MB/s
USB 3.0 ASMedia (back): W/R: 10 MB/s / 125 MB/s
USB 3.1 ASMedia (back): W/R: 10 MB/s / 145 MB/s
*) On the first try file was written at 6.5 MB/s, just like for USB 3.1 speed from my first USB speed test - maybe it is some Windows internal file cache problem?
I know that my way of testing is not the perfect one (I have not found any other one which could work with pendrive) but installation of Switch Driver on "PCI Bus" helped a lot with read speeds, so it seems to me that presence (or absence) of "PCI Express Root Complex" named device should not determine the installation (or not) of Switch Driver.
Two additional notes:
- There is a screenshot in my link showing "Device lower filters" - there was no such category before Switch Driver installation - it appeared just after.
- Before Switch Driver installation I had some "weird problem" with my older 2.0 pendrives - some of them have LED light to show if the pendrive is active or not. After "safely eject pendrive" the light still stayed "ON" till I physically unplug the pendrive. Then i found this link: https://support.microsoft.com/en-us/help…oved-usb-device and edited registry by adding DisableOnSoftRemove - it solved my "light" problems - as soon as I "safely eject pendrive" the light was going off. Now, the "funny thing" is: before Switch Driver installation I removed DisableOnSoftRemove from registry to test this out - it turned out that Switch Driver installation removed the need for DisableOnSoftRemove - now light is going off without this registry edit.
@Lex :
Thank you very much for having taken the time for testing intensively the function and usefulness of the Intel “USB Switch Driver” and for your interesting report.
As I have already written, any system may be different regarding the various USB 2.0/3.0/3.1 manufacturers, device models and their dependencies. That makes it rather difficult to write a short guide, which will be valid for everyone.
According to my experience the best way to get the “pure” Intel USB 3.0/3.1 drivers completely and properly installed is to use the DPInst option. Thie DPInst Utility will find the correct Intel “Switch Device” much easier and safer than the “normal” user.
If you have a specific idea about how I should update the start post of this thread, please let me know it.
@Fernando
I have update Post 1453# with a comparative benchs between old/new ASMedia USB 3.1 driver versions.
It does seem for USB2 and USB3 devices there are no significant differences.
However, the performance is far better for the old version when a USB3.1 device is attached on the ASMedia 3.1 port (example is with an external USB3.1 device/SATA6G closure for a 850EVO SATA6G SSD).
@100PIER :
Thanks for having done the additional benchmark comparison tests and for the update of >this< post.
So I will leave the start post regarding the offered ASMedia USB 3.0/3.1 drivers as it is now.
Yes, I think so.
About the TRIM not working is it normal when an external USB3.1 closure is used for a SATA6G SSD device (such as 850EVO) ?
TRIM obviously doesn’t pass through any USB Controller.
