ROG STRIX X470-I USB 3.1 Gen2 issues

[aaraujo]

I am using a StarTech.com USB312SAT3CB SATA III to USB 3.1 Gen2 (10 Mbps) adapter
https://www.startech.com/HDD/Adapters/usb-3-1-gen-2-sata-adapter-cable~USB312SAT3CB
with a Samsung SSD 850 PRO
https://www.samsung.com/semiconductor/minisite/ssd/product/consumer/850pro/

The motherboards is an ASUS ROG STRIX X470-I GAMING. I am having issues getting one of the USB 3.1 red ports at the back to work with this SATA to USB 3.1 Gen2 adapter. One of the USB 3.1 red ports works perfectly fine. However the other USB 3.1 red port detects the device at USB 3.1, but the drive disappears a few seconds afterwards, and can never be used with this port. The operating system is Windows 10 64 bit. I have tried the same USB 3.1 port under Linux but the drive never appears. Again the other of the two USB 3.1 red ports works fine under Linux.

I am posting here trying to know if somebody else with this motherboard is having issues getting the red USB 3.1 ports to work. I have tried to attach a USB 3.0 Flash drive to the same USB 3.1 red port, but the drive works perfectly fine! The problem arises when a USB 3.1 device is connected to that port.

I am also having VERY VERY SLOW speed transfers between the same USB 3.1 red ports at the back, and a LG G6 smartphone. I am using an Anker USB-A to USB-C 3.1 Gen2 certified cable. The same SLOW speeds are observed while using the USB 3.0 ports at the front. I get normal speeds while using the blue USB 3.0 ports at the back. The cable is this one
https://www.anker.com/products/variant/anker-powerline-ii-usb-c-to-usb-3-1-cable-3ft/A8465011

Is there anyone with an ASUS ROG STRIX X470-I motherboard who can check if the USB ports are working and report here? Thank you!

[Stellarspace]

1. Do you have the latest BIOS and drivers?
https://www.asus.com/us/Motherboards/ROG-STRIX-X470-I-GAMING/HelpDesk_Download/
I suggest flashing the BIOS with the latest version and installing the latest USB drivers for your motherboard.

2. According to an article I read a couple days ago, the USB 3.1 standard varies between manufacturers. Just because it's labeled as 3.1 does not mean it supports 10 Gbps speeds. Unfortunately, there isn't really a standard.

3. Have you used https://crystalmark.info/en/software/crystaldiskinfo/ to test the speed on the 2 x USB 3.1 Gen 2 port(s) (2 at back panel, red, Type-A)?
I would like to point out that actual speeds vary depending on what data is being transfered and how high the read/write speeds are.

[aaraujo]

1. Yes, I am using the latest BIOS which is 0601. The motherboard arrived with that BIOS version from the store. I am using USB 3.1 drivers from Microsoft under Windows 10. I cannot find USB 3.1 drivers for Windows 10 neither on AMD driver website nor on the ASUS driver website.

2. I can confirm that the motherboard indeed works at more than 5 Gbps on one of those two red ports. As I wrote above one of the USB 3.1 red ports works just fine. I tested under CristalDiskMark. Otherwise how could I get 550 MB/s read and write speeds under that port. On regular USB 3.0 port I get around 450 MB/s speed. Also AIDA64 informs that the device gets detected as USB 3.1 on those two red ports. The problem is that one of the red ports is not working properly! The other is working perfectly fine. I don't believe this is an issue with the standard.

3. I have checked with AIDA64, which tells me that the device is working as USB 3.1 on those two ports. The problem is not the speed. The problem is that the drive disappears when using one of those two red ports. I cannot do nothing with that SSD on that red port, much less check the actual speed!

I just would like to get confirmation from someone using this very same motherboard. If everybody else is observing the same issue there is no point in RMAing the board. In that case it is likely a driver or BIOS issue.

I mentioned that the same issue is observed under Linux. It is very unlikely that both Windows and Linux have the same driver bug, so I raise two possibilities, either the BIOS is buggy, or there is a hardware defect with my sample. I need to know which of these is the culprit before doing anything else.

[aaraujo]

I was using "level 5" at the soc load line calibration. When I switched to "level 3" I noticed that I could use the USB 3.1 drive on the problematic port. Still CrystalDiskMark could not complete the benchmark before the drive disappearing from Windows. I ran memtest and got errors on test 8. I believe these memory errors are related to the issue. I tried "level 2" soc load line calibration and memtest completed 12 passes of test 8 without any error. I have to do more testing. But I believe I am in the right path to find the solution to the issue.

[aaraujo]

I tried running memtest in order to check the memory for possible problems. I ran memtest for 18 hours and no error was found. I also ran prime95 for more than 12 hours without finding any error! The memory and the CPU are fine.
A new BIOS version is out. So I flashed BIOS version 0804 to the motherboard. Now I still have the issue with that USB port! The other USB ports are working fine.
I contacted ASUS, who suggested I should contact the store for an RMA. I contacted the store and I am still waiting for their reply.

[aaraujo]

UPDATE:

Tried another ASUS X470-I motherboard, and I am experiencing the same issues with this unit. I noticed that the USB 3.1 gen2 port which works fine, is closer to the board, while the one which does not, is the one on the top, which lies more distanced from the board. This leads me to hypothesize that this is a design flaw. The same wire gauge is being used to transmit the signal to both USB ports. This explains why the port located farther apart from the board experiences slow transfer speeds and random disconnects. A back panel with the correct wire gauge would fix this issue. The top port needs a wire with a larger diameter to carry the electrical signal. This is a physical issue, and a BIOS update is very unlikely to fix it.

BTW, the same USB 3.1 gen2 to SATA adapter and SSD works perfectly fine on all the USB 3.1 gen 1 (blue ports). There are 4 USB blue ports on the back. There are 2 USB blue ports at mid board, which I use wired to the front part of the case. These blue ports work fine with this adapter, but at USB 3.0 speeds. This is expected, so is perfectly fine.

The PSU is a Seasonic PRIME Titanium Fanless:
https://seasonic.com/prime-titanium-fanless

It registers around 4.960 Volts on the +5V rail. This is the rail that supplies current to USB. It can supply enough current to power around 22 USB 3.0 ports drawing 900 mA, the maximum current according to spec. This motherboard has 8 USB ports, and I am not using all of them. I am only using a single NVME drive and a graphics card. The graphics card and the CPU get current from the +12V rail. I am not sure about the NVME drive. I do not think the PSU is the cause of these issues.

[aaraujo]

Found this related issue:
https://www.techpowerup.com/223583/amd-zen-processor-integrated-chipset-has-usb-3-1-issues-could-esc...

[aaraujo]

Update.

Quote from this ASUS ROG Strix B450-I Gaming Motherboard review:

https://www.kitguru.net/components/motherboard/ryan-martin/asus-rog-strix-b450-i-gaming-motherboard-...

USB 3.0 performance was as expected while USB 3.1 performance experienced numerous performance and compatibility issues similar to those we observed with the ASRock X470 Gaming-ITX/ac motherboard.

Effectively, there’s a bug where connected USB 3.1 devices disconnect during heavy and sustained loading. We experienced this bug using a USB 3.1 ASM1352R RAID box and ASM1351 USB 3.1 to SATA III device with the OCZ Trio 150 480GB SSD.

The only way we were able to complete an ATTO run from start to finish without interruption while using USB 3.1 was with the ASM1351 USB 3.1 to SATA III device and a Transcend SSD370S 512GB SSD.

Given the MSI B450 Tomahawk successfully completed USB 3.1 testing using the B450 chipset controller we can only assume this is a motherboard specific problem. It’s difficult to say which USB 3.1 component causes the problem, as it is unlikely to be the controller.

We expect this ASUS motherboard and the ASRock X470 Gaming-ITX/ac must have a similarity somewhere as both failed in similar ways when using USB 3.1, our best educated guess at this stage is that it is potentially a compatibility issue with USB 3.1 redriver chip.

[aaraujo]

Update:

I tried a different board, this time an ASUS ROG STRIX X370-I GAMING

With Microsoft's USB 3.1 drivers installed, USB 3.1 slows down after a while.
With AMD USB 3.1 drivers installed, both USB 3.1 ports work as they should. I am able to complete all the tests.

I confirm that ASUS ROG STRIX X470-I GAMING has USB 3.1 issues with both Microsoft and AMD USB 3.1 drivers installed under Windows 10.

To install AMD USB 3.1 drivers under Windows 10, these websites can be followed:
https://www.reddit.com/r/Amd/comments/6q8qv7/update_regarding_my_usb_issues_with_ryzen/
https://community.amd.com/thread/220104

These boards have been confirmed to have USB 3.1 issues:
ASUS ROG STRIX X470-I GAMING
ASUS ROG STRIX B450-I GAMING
ASRock X470 Gaming-ITX/ac

USB 3.1 issues have been reported with the last two boards by kitguru:
https://www.kitguru.net/components/motherboard/ryan-martin/asus-rog-strix-b450-i-gaming-motherboard-...

I confirm that ASUS ROG STRIX X370-I GAMING works with AMD USB 3.1 drivers under Windows 10.

[aaraujo]

Update:

ASUS ROG STRIX X370-I GAMING

Windows 10 with Microsoft USB 3.1 drivers -> Issues.
Windows 10 with AMD USB 3.1 drivers -> There are no issues!
Fedora Linux 29 -> Issues.

ASUS ROG STRIX X470-I GAMING

Windows 10 with Microsoft USB 3.1 drivers -> Issues.
Windows 10 with AMD USB 3.1 drivers -> Issues.
Fedora Linux 29 -> Issues.

This leads me to conclude that there is no proper device driver support for the USB Redriver chip which is needed to transmit the signal to the more distanced port. Only the ROG STRIX X370-I GAMING works under Windows with the AMD 3.1 device drivers designed for Windows 7. None of the two boards works properly under Linux. AMD did not update the USB drivers for over than a year now. I already RMAed the board twice. It is very silly the way manufacturers deal with these issues!