Ryzen 3900X USB Reset Problem

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
I've been running Linux on a couple of 3600Xs @4.3ghz 24/7 since October of last year and they've been working perfectly. I was so impressed by them that I decided to upgrade one to 3900X since the 5000 series is hard to get and not worth the cost.

I upgraded one of the machines and everything seems OK. I even needed less voltage to get a stable 4.3ghz OC. But after a while I noticed that USB ports would reset at random times and sometimes excessively.

Code:
[Sat Jan 16 06:51:18 2021] usb 4-1: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 07:18:25 2021] usb 4-1: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd

Even though I was super careful, I thought I damaged the motherboard somehow. So I put the 3600X back and it worked perfectly with no resets. Swapped the 3900X back in and the USB resets came back. I tried it in a Supermicro case with a 1200W platinum power supply, but that didn't fix it. I downclocked the CPU to 4.2, 4.1, and finally 4.0ghz, but that didn't fix it. Cleared the CMOS and set up again, but that didn't help either. Ran it headless with only a Samsung USB stick and no other hardware attached; the problem still occurred.

I found this Reddit post (USB problems on B550) with some people having this problem, but with a 5000 series. I'm using a brand new 3900X that came direct from AMD and an MSI B550M Mortar.

Does anyone have any suggestions? I can't seem to find were c-states are in the BIOS.
 
Last edited:

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
I updated to the latest BIOS, set the CPU and RAM frequencies and voltages. Plugged in a single USB stick and keyboard and still getting resets. This time I tried the Gen 2 ports. So it looks like this problem affects all of the ports.

Code:
[Sat Jan 16 11:52:45 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 11:53:39 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 11:57:31 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 11:59:45 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:00:29 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:03:27 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:03:35 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:04:08 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:04:13 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:06:17 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:08:15 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:10:52 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:13:07 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:13:24 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:19:51 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:21:43 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:22:54 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:23:01 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:26:03 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:26:41 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:28:26 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:28:33 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:31:39 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:31:40 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
[Sat Jan 16 12:33:03 2021] usb 4-4: reset SuperSpeed Gen 1 USB device number 2 using xhci_hcd
 

schizo

[H]ard|Gawd
Joined
Nov 6, 2004
Messages
1,654
I also read about problems with USB ports with PCIe 4.0 devices on Zen 3. The fix in those cases was to set the GPU to PCIe 3.0. If that doesn't apply to you, I'm pretty much stumped.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
I also read about problems with USB ports with PCIe 4.0 devices on Zen 3. The fix in those cases was to set the GPU to PCIe 3.0. If that doesn't apply to you, I'm pretty much stumped.

That's the thing, it's a headless server with no GPU installed. I'll put a GPU in, enter the BIOS, and force PCIe to Gen3 where ever possible.

I think the problem occurs when the CPU is loaded. I just ran it for a while unloaded and there were no USB resets. As soon as I put a 100% load on the CPU the USB started to reset.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
It's definitely a problem when the CPU is loaded. I will try disabling some core to see if that helps. I don't know if there is a motherboard problem or I got a defective CPU.

Sensor data at 100% load. Voltages look okay to me.
Code:
k10temp-pci-00c3
Adapter: PCI adapter
Tctl:         +63.5°C  
Tdie:         +63.5°C  
Tccd1:        +63.2°C  
Tccd2:        +61.0°C  

nct6687-isa-0a20
Adapter: ISA adapter
+12V:           12.10 V  (min = +12.10 V, max = +12.24 V)
+5V:             5.11 V  (min =  +5.11 V, max =  +5.14 V)
+3.3V:           3.36 V  (min =  +0.00 V, max =  +3.36 V)
CPU Soc:         1.08 V  (min =  +1.07 V, max =  +1.09 V)
CPU Vcore:       1.29 V  (min =  +1.28 V, max =  +1.29 V)
CPU 1P8:         1.84 V  (min =  +1.84 V, max =  +1.84 V)
CPU VDDP:        0.00 V  (min =  +0.00 V, max =  +0.00 V)
DRAM:            1.41 V  (min =  +1.41 V, max =  +1.41 V)
Chipset:         1.09 V  (min =  +1.09 V, max =  +1.20 V)
CPU Fan:       2090 RPM  (min = 1168 RPM, max = 2090 RPM)
Pump Fan:         0 RPM  (min =    0 RPM, max =    0 RPM)
System Fan #1:    0 RPM  (min =    0 RPM, max =    0 RPM)
System Fan #2:    0 RPM  (min =    0 RPM, max =    0 RPM)
System Fan #3:    0 RPM  (min =    0 RPM, max =    0 RPM)
System Fan #4:    0 RPM  (min =    0 RPM, max =    0 RPM)
System Fan #5:    0 RPM  (min =    0 RPM, max =    0 RPM)
System Fan #6:    0 RPM  (min =    0 RPM, max =    0 RPM)
CPU:            +63.0°C  (low  = +22.0°C, high = +63.0°C)
System:         +24.0°C  (low  = +20.0°C, high = +24.0°C)
VRM MOS:        +34.0°C  (low  = +18.0°C, high = +34.0°C)
PCH:            +23.0°C  (low  = +17.0°C, high = +23.0°C)
CPU Socket:     +39.0°C  (low  = +16.0°C, high = +39.0°C)
PCIe x1:        +21.0°C  (low  = +15.0°C, high = +21.0°C)
M2_1:            +0.0°C  (low  =  +0.0°C, high =  +0.0°C)
 

TheSlySyl

[H]ard|Gawd
Joined
May 30, 2018
Messages
1,047
Do you have usb selected suspend on in the bios or power options?
I've never came across anything like this with my 3900X.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
Do you have usb selected suspend on in the bios or power options?
I've never came across anything like this with my 3900X.

Hi. Thanks for the suggestion. I'm using an MSI B550M Mortar and it doesn't have that option anywhere in the BIOS.
 

TheSlySyl

[H]ard|Gawd
Joined
May 30, 2018
Messages
1,047
I just know i've had issues in the past with motherboards turning off usb ports due to some power saving feature, so that might have been something to look into.
 

Dan_D

Extremely [H]
Joined
Feb 9, 2002
Messages
57,767
Keep in mind, the Ryzen 9 3900X has a USB controller integrated into it. There could be an issue with that controller. Make sure your Infinity Fabric clock isn't set too high either. That can cause all sorts of issues.

1610833302080.png
 

schizo

[H]ard|Gawd
Joined
Nov 6, 2004
Messages
1,654
Yeah, see if it still happens after resetting your BIOS to defaults, and if so, if adding voltage fixes it. May be a poor sample on your CPU.

Also for your headless home servers, have you heard of PiKVM? I just found out about this project myself. It's badass, an IP KVM for <$100.

https://github.com/pikvm/pikvm
 

KarVi71

n00b
Joined
Dec 14, 2016
Messages
32
I dont know if this applies to your situation, but back when I had a 1700 on a B350, I had all kinds of weird USB issues.

I found that increasing the Soc voltage 0.05V (or perhaps it was 0.1V) via offset, removed the problems.

The 3900x is a completely different CPU, so I don't know if it will help. But its worth a shot.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
Dan, I suspected that it might have the USB built it, but I wasn't sure. Make sense. But aren't these CPUs guaranteed to work at 1800Mhz Infinity Fabric?

Schizo, I'm running everything at default now (2666Mhz RAM and 1333Mhz IF) and I thought everything was working. But it took a little over 50 minutes and I got a USB reset. I think there is a bit of randomness to it. It will probably act up again later.

KarVi71, thanks for the suggestion. I'll try bumping the SOC voltage to see if it helps. I'll also try one of the SOC load line calibration options to see if that helps.

This sucks; this CPU doesn't work properly at stock speeds and voltages.
 

loafer87gt

Limp Gawd
Joined
Feb 2, 2005
Messages
435
Dan, I suspected that it might have the USB built it, but I wasn't sure. Make sense. But aren't these CPUs guaranteed to work at 1800Mhz Infinity Fabric?

Schizo, I'm running everything at default now (2666Mhz RAM and 1333Mhz IF) and I thought everything was working. But it took a little over 50 minutes and I got a USB reset. I think there is a bit of randomness to it. It will probably act up again later.

KarVi71, thanks for the suggestion. I'll try bumping the SOC voltage to see if it helps. I'll also try one of the SOC load line calibration options to see if that helps.

This sucks; this CPU doesn't work properly at stock speeds and voltages.
I am having the same issue with my 3950x / X570 combo. Here is my thread on the issue. I'll let you know if I find a fix.

https://hardforum.com/threads/gigab...usb-devices-disconnecting-under-load.2006790/
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
I am having the same issue with my 3950x / X570 combo. Here is my thread on the issue. I'll let you know if I find a fix.

https://hardforum.com/threads/gigab...usb-devices-disconnecting-under-load.2006790/
Absolutely nothing you do will fix this problem. The best I've gotten was get the USB 2.0 (chipset ports) working okay by setting the chipset PCIe to Gen 2. People who say they've fixed the problem by disabling C-states, setting idle power, disabling USB sleep, etc, etc haven't loaded their CPUs 100% for more than five minutes and watched their USB devices stutter and disconnect.

I have wasted too much time and money on this buying and testing different motherboards, RAM, and trying out three different PSUs. It's a shame because the CPU I have is a 24 karat gold overclocker. But without USB working, the system is nearly useless.

My opinion is that AMD is selling defective chips and hoping people don't notice or pass the blame unto others.

I'll have to settle for a 3600X that works. The 3900X is going back.
 

Kubicide

Limp Gawd
Joined
Oct 15, 2005
Messages
134
Hmm. I doubt it's the CPU. It's likely MB related and the 3900X is somehow unmasking this problem. I've ran a 3900X and now a 5900X on a Gigabyte board that's on ~8-10 hrs daily for the past year, with all USB devices populated, and have had zero issues. And it's not a common problem or not even reported on many other MB's.

Note you're not alone as other users have same issue but with different CPUs:

USB 2 disconnections on B550M MORTAR and Ryzen 5 5600x (Zen 3) | MSI Global English Forum - Index
Having constant issues with MAG B550M Mortar | MSI Global English Forum - Index



Edit: should add it also may very well be s/w related such as drivers or even a specific Win 10 version. Check your USB drivers and also check for the latest chipset drivers from AMD.
 
Last edited:

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
Hmm. I doubt it's the CPU. It's likely MB related and the 3900X is somehow unmasking this problem. I've ran a 3900X and now a 5900X on a Gigabyte board, with all USB devices populated, and have had zero issues. And it's not a common problem or not even reported on many other MB's.

Note you're not alone as other users have same issue but with different CPUs:

USB 2 disconnections on B550M MORTAR and Ryzen 5 5600x (Zen 3) | MSI Global English Forum - Index
Having constant issues with MAG B550M Mortar | MSI Global English Forum - Index



Edit: should add it also may very well be s/w related such as drivers or even a specific Win 10 version. Check your USB drivers and also check for the latest chipset drivers from AMD.
It's the CPUs. I tried: MSI B550M Mortar (two of them), MSI B550M PRO VDH, and an ASUS X570-Plus TUF. All have the exact same problem.

It's also not software, drivers, or OS. Linux and Windows both have the same problem.

You're just lucky you didn't get a defective CPU.
 

loafer87gt

Limp Gawd
Joined
Feb 2, 2005
Messages
435
Absolutely nothing you do will fix this problem. The best I've gotten was get the USB 2.0 (chipset ports) working okay by setting the chipset PCIe to Gen 2. People who say they've fixed the problem by disabling C-states, setting idle power, disabling USB sleep, etc, etc haven't loaded their CPUs 100% for more than five minutes and watched their USB devices stutter and disconnect.

I have wasted too much time and money on this buying and testing different motherboards, RAM, and trying out three different PSUs. It's a shame because the CPU I have is a 24 karat gold overclocker. But without USB working, the system is nearly useless.

My opinion is that AMD is selling defective chips and hoping people don't notice or pass the blame unto others.

I'll have to settle for a 3600X that works. The 3900X is going back.
It's very frustrating indeed. What I find very strange is I can load my CPU using OCCT and P95 and its fine, but I have some 3DSMax scenes that cause the USB disconnects instantly on render. I did the BIOS update, installed the latest chipset drivers, did the other suggested BIOS changes and the problem still persists. One thing I was surprised to see is that at auto my SOC voltage was set to 1.25 when most guides say 1.1-1.2 is a safe range. I wonder if it is because of the 4 DIMMS I am running?

Edit: One thing I noticed is that even though this I have my SOC voltage set to Auto (1.25 in BIOS), using HWInfo it shows as 1.087 and drops as low as 1.06 when under load. I wonder if this could be cause for the dropping devices?
 
Last edited:

x509

2[H]4U
Joined
Sep 20, 2009
Messages
2,390
t's a shame because the CPU I have is a 24 karat gold overclocker. But without USB working, the system is nearly useless.
I also have a 3900X (ASUS Strix-E motherboard) which I have not yet tried to OC. But I keep reading that with PBO there is much less reason to do manual settings to OC the 3000 series CPUs. Was I reading the wrong posts?
 

loafer87gt

Limp Gawd
Joined
Feb 2, 2005
Messages
435
I also have a 3900X (ASUS Strix-E motherboard) which I have not yet tried to OC. But I keep reading that with PBO there is much less reason to do manual settings to OC the 3000 series CPUs. Was I reading the wrong posts?
I have not done any manual overclocking with my 3950X and am running into these issues. I did try manually setting my SOC voltage to 1.2 using Ryzen Master and it was reading as 1.197 using HWInfo up from the 1.087 using the auto voltage. I still get the USB disconnects when full load is applied to the computer though. I'm at a total loss where to go from here.
 
Last edited:

Kubicide

Limp Gawd
Joined
Oct 15, 2005
Messages
134
It's the CPUs. I tried: MSI B550M Mortar (two of them), MSI B550M PRO VDH, and an ASUS X570-Plus TUF. All have the exact same problem.

It's also not software, drivers, or OS. Linux and Windows both have the same problem.

You're just lucky you didn't get a defective CPU.
I still highly doubt it's the CPU. Intel users have reported the same thing of USB disconnects galore. Most often then not it's drivers and/or BIOS settings/problems and/or conflicts from other bad behaving USB devices/drivers.

BTW do you have any BT devices attached? Or have any adapters like USB to SATA? You keep coming back with additional detail at times so I think we're not seeing the full story with your setup. So it's hard to help if there's potentially important info being held back.
 
Last edited:

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
I still highly doubt it's the CPU. Intel users have reported the same thing of USB disconnects galore. Most often then not it's drivers and/or BIOS settings/problems and/or conflicts from other bad behaving USB devices/drivers.

BTW do you have any BT devices attached? Or have any adapters like USB to SATA? You keep coming back with additional detail at times so I think we're not seeing the full story with your setup. So it's hard to help if there's potentially important info being held back.

More details? Exactly what makes you suspicious that I might be lying or intentionally omitting details? Maybe there are more details because I have wasted more time and money trying to solve this. And now I'm wasting more time replying to your posts.

Setup: Case, PSU, motherboard, 3900X, RAM, USB stick, Linux, software, nothing else. No video card, no SATA, no SSD., nothing else.

Some people have problems with USB and they are using 3600Xs. Both my 3600Xs work perfectly. There is nothing but a thumb drive attached during testing and it disconnects, not even a video card. I've tried two brand new Samsung Fit and a PNY Attaché USB 3 thumb drive. Oh no I added more detail!

Did you read that I'm sending the CPU back? You can infer from that that I'm not asking for help on this any longer. And did you bother to read this and the comments? https://www.reddit.com/r/Amd/comments/kbr6zl/usb_issues_on_b550x570_motherboards/?sort=new

From my testing, everything points to the CPU, not the motherboard, as I have two 3600Xs that work perfectly and I have tested three MSI boards (my two B550 Mortars and a Pro VDH) and one ASUS board with the 3900X and the same USB problem occurs on all of them.

I've returned the MSI Pro VDH to Amazon and I got the ASUS board packed and ready for UPS to pick it up Monday. Waiting for AMD to reply so I can ship the 3900X back. Maybe they'll send me another CPU to test, but I'd rather just stick to the 3600Xs and the MSI Mortars that work perfectly together.

Again, I am no longer asking for help or advice. But I hope this thread helps people who have USB problems on their AMD 3000/5000 setups. My advice: don't waste your time and money. Just return your new CPU and motherboard and wait for AMD to fix the problem and then buy again.
 

loafer87gt

Limp Gawd
Joined
Feb 2, 2005
Messages
435
Thanks for the update, Dr. Nut. I just recieved my externally powered USB hub, and even with this connected by devices still drop out under full load. I don't know where to go from here. AMD and Gigabyte both say my external USB devices are faulty, but I know this is not the case as they all work fine on my other computers.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
Thanks for the update, Dr. Nut. I just recieved my externally powered USB hub, and even with this connected by devices still drop out under full load. I don't know where to go from here. AMD and Gigabyte both say my external USB devices are faulty, but I know this is not the case as they all work fine on my other computers.
What I've gathered is that only the USB ports that come from my 3900X are the ones that are seriously affected. Yet the CPU USB ports on my 3600X on the same motherboard work fine. This is why I am fairly confident the 3900X is defective (yours probably is too.) I have my USB OS/boot thumb drive on the chipset USB 2.0 ports and it is working fine now. I have yet to try the chipset USB 3.x ports, but I'll assume they are working fine.

If you want to live with it, you could try moving all of your USB devices to the chipset ports and get a PCIe USB 3.x card if you need more ports. You could also press AMD for an RMA. Or try to exchange it from where ever you bought it.

I'm still waiting for AMD's response.
 

schizo

[H]ard|Gawd
Joined
Nov 6, 2004
Messages
1,654
Unshielded USB3 can interfere with 2.4Ghz wireless signals, that's very common.

The USB reset thing on newer AMD AGESA versions is something different, that's just a bug. Most common fix I've seen is to disable PCIe 4.0.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
Unshielded USB3 can interfere with 2.4Ghz wireless signals, that's very common.

The USB reset thing on newer AMD AGESA versions is something different, that's just a bug. Most common fix I've seen is to disable PCIe 4.0.

No one is talking about WiFi interference here. Why are you even bringing that up? It makes no sense.

This is not something new with AGESA. It's not a software bug. Some Ryzens have defective USB. I have a defective 3900X. I swap a good 3600X and the on CPU USB works perfectly. None of those anecdotal disable this or that "fixes" work.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
To be 100% sure about the CPU, I swapped the 3900X into my other B550 Mortar that has been running perfectly for months. I also used a Hyper 212 cooler instead of the eSports Duo. Of course it didn't help. The resets began around 34 minutes after putting a 100% load on the CPU.

So either I got four defective motherboards, or AMD sent me a bad CPU. Oh, and the B550M Mortars have the latest BIOS with AGESA 1.2.0.0. I know some people were thinking it would fix the problem, but it doesn't.

I received a reply from AMD today, and of course they passed the blame off to the motherboard manufacturer and completely ignored the fact that I requested a return. For such a large company, AMD's store and tech support are abysmal. Let's see if they ignore my second request to return and get a refund.
 

schizo

[H]ard|Gawd
Joined
Nov 6, 2004
Messages
1,654
Problems can result from different root causes. Disabling PCIe 4.0 does work for many people. If you have a defective CPU or motherboard, then sure, that can do it too.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
Update:
I had to go to call my CC bank and initiate a dispute so that I can get my money back for the defective CPU. A few days after that, AMD's store (apparently run by Digital River) caved and sent me a return label. I would suggest not buying CPUs and GPUs from AMD's online store. I think Digital River also runs NVIDIA's store and maybe Microsoft's. So avoid those too.

In the mean time, I managed to borrow someone's 3700X and try it out on one of my B550 Mortars and the USB ports tied to the CPU worked perfectly. I bought a plain 3600 and tested it on the B550 Mortar; works perfectly. Maybe Ryzens with multiple CCDs are more prone to this problem.

One thing I did notice is that USB ports coming from the B550 chipset sometimes reset/disconnect regardless of which CPU model is used. So not only is AMD selling defective CPUs, but also defective chipsets. Fortunately the chipset USB reset/disconnect problem seldom happens (about once every few days) and it always reconnects, so I can live with that.

I'm going to buy another 3900X and hope that the USB ports from the CPU work properly.
 

schizo

[H]ard|Gawd
Joined
Nov 6, 2004
Messages
1,654
There was a post on the AMD subreddit indicating AMD is aware of USB issues on 550/570 platforms and trying to replicate. So it's probably a motherboard, BIOS, or AGESA issue, since it happens with both Zen2 and 3. Nobody knows what triggers it, I haven't had the problem on my Asus B550-F board but others have.
 

Dr.Nut

Limp Gawd
Joined
Dec 13, 2011
Messages
171
There was a post on the AMD subreddit indicating AMD is aware of USB issues on 550/570 platforms and trying to replicate. So it's probably a motherboard, BIOS, or AGESA issue, since it happens with both Zen2 and 3. Nobody knows what triggers it, I haven't had the problem on my Asus B550-F board but others have.
I pored over Reddit trying to find a solution and found nothing conclusive. Not only that, but I can't find anything as to AMD confirming or admitting the problem. If you have post that says all of that, then link to it here, otherwise it did not happen. I have spoken to someone at AMD tech support and they said nothing, only took my information and nothing else. I got an email from AMD passing the fault on my motherboard and to contact the motherboard manufacturer. I told the techs I tried various boards three MSI and one ASUS.

Here's AMD's response:
Response and Service Request History:

Thank you for your email.

I understand your concern that USB ports are not functioning but they are not part of your processor, kindly contact your motherboard manufacturer.

In order to update this service request, please respond without deleting or modifying the service request reference number in the email subject or in the email correspondence below.

He doesn't even know that the CPU has four 10G USB PHYs coming from its IOD. Serously. Total BS out of AMD.

I've tested various CPUS and various motherboards. Again, it's not the motherboard that has problems, in my case I am 100% confident the CPU was faulty. Only the 3900X I bought had flaky USB. Maybe I was really unlucky to get a dud and this problem is not endemic to AMD's 3000 and 5000 series CPUs. On the other hand, maybe AMD has a bug in its IOD. I'll get another 3900X soon and seen how that goes.
 

learners permit

Limp Gawd
Joined
Jun 15, 2005
Messages
499
Couple years ago I had a similar issue on an Intel X99 rig. Very irritating I know. I just figured the board was getting old and the usb hubs were crappin out so I built a ryzen system on X570 and a 3800x. A few days later the usb hub started doing the same thing on the new rig and I was thinking wth is going on. Long story short it turned out that my trackball mouse used on both systems was intermittently overloading the usb hubs power capability and shutting it down via overload protection in the hardware. First time I had ever seen that happen with any component. The mouse gave no other indication of failure as it still worked without issue other than the hub shutting off randomly.
 

x509

2[H]4U
Joined
Sep 20, 2009
Messages
2,390
Couple years ago I had a similar issue on an Intel X99 rig. Very irritating I know. I just figured the board was getting old and the usb hubs were crappin out so I built a ryzen system on X570 and a 3800x. A few days later the usb hub started doing the same thing on the new rig and I was thinking wth is going on. Long story short it turned out that my trackball mouse used on both systems was intermittently overloading the usb hubs power capability and shutting it down via overload protection in the hardware. First time I had ever seen that happen with any component. The mouse gave no other indication of failure as it still worked without issue other than the hub shutting off randomly.
learners permit How did you figure out the problem with the trackball mouse?
 
Top