General Discussions

[-sum1-]

I realise I made the wrong call by posting this issue to that annoying place called Reddit 6 months ago, so I'm posting it again here where it belongs.

Hi.

I've made this topic because I've discovered I have an issue which has terrible coverage online and no single hub of discussion. I only discovered its cause 7 months after I first had it, because it was notoriously difficult to reproduce.

According to my research, this issue affects AMD CPUs (at least Ryzen, FX and Phenom). Motherboard doesn't matter, Windows version doesn't matter, graphics card doesn't matter, RAM doesn't matter.

Issue:

At random times, the monitor(s) will cut to a black screen (no signal), and fans will immediately speed up to default-default settings (pre-POST). Power continues to be supplied (including to peripherals), but nothing will work, not even the power button or reset switch, forcing you to switch off the PSU or mains power and switch it on again. For me, half the time, the PC would automatically reboot after the blackscreen, exactly mimicking the effect of the reset button (though I ruled out the reset button as a cause by unplugging it). I did not have any BSODs or minidumps. There was nothing about it in my Event Viewer either.

The frequency of the issue is very unpredictable, but it started immediately after I built my computer. It happened to me on the following dates (I kept a manual log):

• 31 August 2018 and 2, 10, 17, 18 September 2018

• 8, 9, 10 February 2019

• 8, 9, 10, 13, 14, 15 March 2019

HWiNFO logs leading up to the crashes in March showed nothing out of the ordinary, even in the last second before blackscreen. Stress testing in Prime95 and Superposition could not induce the issue. On one day in September, the computer even woke up from sleep straight into this black screen state.

Cause:

I noticed during March that the problem was occurring when I had HWiNFO, MSI Afterburner and ASUS AI Suite open at the same time. I then managed to confirm online that the cause of the issue was some sort of conflict involving multiple hardware-monitoring softwares running. Sources specifically point to a CPU thermal sensor issue.

I was not monitoring the HWiNFO ASUS sensors which prompt a warning of instability in March. Also, the problem previously occurred when I had CPU-Z, CoreTemp or Speccy running instead of HWiNFO. The problem may also have happened when only AI Suite and Afterburner were running. Other users online report the issue with various other programs that monitor sensors.

Fix:

The workaround will be obvious – to avoid using multiple programs that could monitor the CPU temp sensor. But I've seen no information about a fix or the issue being resolved by AMD themselves. There are hints in some of the topics below that BIOS or driver updates have reduced the symptoms, but the only surefire way I know of to avoid this issue is to avoid running multiple hardware-monitoring processes (in my case, more than 2, it seems).

External links to topics of this issue:

• https://rog.asus.com/forum/showthread.php?102145-Random-black-screen-on-ROG-STRIX-X370-F-Gaming-(wit...

^ caused by ASUS Aura temperature-based colour mode in a possible conflict with other software

• https://linustechtips.com/main/topic/891427-ryzen-7-black-screen/ and https://www.reddit.com/r/AMDHelp/comments/804fv7/ryzen_7_random_black_screen/

^ caused by ASUS Aura + ASUS AI Suite

• https://community.amd.com/thread/218356

^ caused by HWiNFO + CPU-Z + Afterburner, or HWiNFO + Corsair Link + Afterburner

• http://forum.corsair.com/forums/showthread.php?t=178422

^ caused by HWiNFO + Corsair iCUE (+ maybe another application)

• https://forums.tomshardware.com/threads/pc-crashes-monitor-losses-signal-pc-stays-on-power-and-reset...

^ caused by AMD Overdrive thermal sensing in a possible conflict with other software

• https://forums.tomshardware.com/threads/amd-ryzen-build-black-screen-no-hdd-lights-power-reset-not-r...

^ caused by AMD Ryzen Master + ASUS AI Suite

• https://linustechtips.com/main/topic/900317-can-someone-weigh-in-on-this-weird-mystery-issue/

^ cause unreported

• https://forums.tomshardware.com/threads/computer-crashes-when-gaming-no-signal-from-monitor-no-sound...

^ cause unreported

[misterj]

As usual this is a self inflicted problem and definitely NOT an AMD problem.  Running junk-ware from MB and other vendors will ultimately be bad news and running multiple ones even more sever problems.  In the case of Ryzen, users should ONLY run Ryzen Master.  Enjoy, John.

[-sum1-]

It's an AMD problem because the only commonality I've seen across every online post that suffers this exact problem is an AMD processor. Also, Ryzen Master is irrelevant, because if you'll check the links I included, Ryzen Master (and AMD Overdrive for that matter) was one of the triggers for the crash, and the problem affects FX and Phenom processors too, at least. You seem to have a history of insisting users use Ryzen Master or screenshot Ryzen Master before helping them. At least read the featured links before wrongly assuming the problem is self-inflicted – the ROG post and others confirm it to be an AMD CPU temp sensor issue, and Intel doesn't seem to behave this way.

I've been running MSI Afterburner and ASUS AI Suite constantly without any issue since the last crashes. Only when I launched a third CPU temp monitoring program like HWiNFO, Speccy or CoreTemp did the computer blackscreen after a few minutes.

[thuum]

Why you would recommend a overclocking tool to pinpoint a systemwide shutdown, is over my head. And why you would consider it less junkwarish, than all the other "junkwares" coming from the manufacturers of specific hardware, even more.

Delving through two threads, it is quite obvious, that AMD does not want you to monitor their hardware...and that is the source of the problem.

[DL_GamZ]

I am having an issue where my screen went black and no signal in the middle of use. I am using all AMD drivers for my AMD RX 6800 and Ryzen 5 CPU. I just updated my Drivers on Monday to the latest Drivers. I tried hard restarting which fixed the issue last time. Than I tried getting Windows to go into Auto Diagnostics mode. Which didn’t work. Is there something else I can do?

[pokester]

Before asking a question the forum headers asks you to provide your system information. I don't really have enough information to offer a suggestion so please give the following information as required to post:

When posting a new question, please provide as much detail as possible describing your issue making sure to include the relevant hardware and software configuration. 
For example:

• Issue Description: [Describe your issue in detail here]
• Hardware: [Describe the make and model of your: Graphics Card, CPU, Motherboard, RAM, PSU, Display(s), etc.]
• Software: [Describe version or release date of your: Operating System, Game/Application, Drivers, etc.]

That being said what you describe sounds more like a power issue and likely affecting GPU not CPU. Once we have specs we can give a better idea of what to try.

As already mentioned AMD does not suggest the use of 3rd party tweaking software such as Afterburner. You can get yourself quickly into an unstable setup doing so.

If you have an AMD GPU you need nothing more than Wattman in the drivers. For your CPU you should use nothing more than Ryzen Master.

[-sum1-]

This is not a question, as referenced in the guidelines Read This First . The post makes it pretty clear what it is, a central topic to centralise discussion about this particular issue. If you read the post and the external links, you'd see that it's confirmed to be a CPU issue, not a graphics card issue (I even included a "Cause" section if you read the post), and by extension it seems that all CPUs reported to be affected by this specific issue are AMD CPUs. I've already solved my problem. This is to help other people who may have been racking their brains for months like I was with the same problem.

[misterj]

-sum1-, I do not in any way believe this is an AMD issue.  It is simple misuse of poor quality utilities.  As pokester said and I have told many users, all that is needed is Ryzen Master.  Enjoy, John.

[pokester]

Well this is for questions for people needing help from other users. You are breaking forum rules then by not posting in general discussion. Your discussion doesn't belong here at all then.I was only trying to get enough information to try to help you.  Nobody can possibly see anything as you didn't give information just a bunch of general comments. I can see that you are pretty rude to people trying to help you. Glad you fixed your issue. Since you still have not really explained your issue with comments spread across literally generations of hardware or what fixed it, I am not sure how your intent is to inform or help the community. Have no fear i am quite sure I nor anyone else will attempt to help you again.  I am sure one of the AMD mods will be happy to put your post where it belongs. amdmatt‌ ray_m‌

[tilt]

pokester‌ I don't find anything rude about his replies and I don't know how he can be more clear about what he was attempting to do. Clearly, he is the one trying to help. I have been searching and searching while testing and testing finding no straight answers over the internet. What I did find was A LOT of people are having this problem with slightly different specs (different mb types, different RAM types, different generation AMD processors) but all are AMD, with no central place out there to DISCUSS the issue. So yes, I get what sum1 was trying to do here.  

Anyway, here is my situation. I changed my i7-7700 cpu on a ASUS prime b250-c motherboard with 16gb hyper X 2400m RAM to AMD 2700x on a ASUS prime x470 pro motherboard. At first i just used the hyper x RAM for couple weeks until I had money to purchase an AMD optimized 3200m RAM having no problems (even used AI suite to OC to 4150 mhz on cpu and still having the monitoring software used on my old intel cpu and board). In the meantime, I went to pc-parts picker and plugged in my build parts to find compatible RAM. I found the Corsair Vengeance LPX  3200M specifically saying AMD optimized right here on the box as opposed to the other suggested compatible RAM not even saying AMD optimized. So for obvious reasons , I went with that.  It was only after I installed this RAM that I began experiencing this exact random blackscreen issue described here.  I firmly believe the RAM change is just a coincidence and surely could not be the issue, but I just reinstalled the old hyper X 2400m RAM today and my system already feels more stable. Is this all in my head? It just doesn't make sense.

[colesdav]

I think the Memory Controller on Ryzen 2000 series and below are "sub optimal"
See: Asus ROG Crosshair Hero VII, Ryzen 2700x, Corsair Vengeance RGB (Pro) 3200 4x16GB Memtest86 failures... 

That link may be of help.

You should have saved yourself a massive headache and just gone for a Ryzen 3000 series processor.

Another thing to improve stability is remove the AMD GPU and drivers and install an Nvidia GPU.

More background to the above link:
https://community.amd.com/message/2952359?commentID=2952359 
Libertas 

[pokester]

Well you are certainly entitled to your opinion. If the OP's information was good for you then that is a good thing. And yes there is proper specific place for this discussion. That would be here. https://community.amd.com/community/support-forums/processors  This is where you would talk about your CPU and Motherboard issues. This wasn't pointed out to badger the OP but for the OP to get traffic from more people likely being able to help the situation. 

[juraj_m_]

I have "ASRock Fatal1ty X470 Gaming K4" and their "F-Stream Tuning" utility (for managing fans, overclocking and monitoring) is also causing this. If I have it running, my PC won't wake up from sleep or hibernation. When I turn on PC, I get black screen and all fans will go full speed (or just loud). However reset (or power) button always worked.

[davegames]

I was having this exact issue.

Random black screens for what appeared to be no apparent reason.

As soonas I turned off Corsair iCue that I had monitoring the CPU temps this stopped occuring.

Really frustrating problem with this being the only source of information on the problem that I could fine.

[colesdav]

Except that my RX Vega 64 / 56 and Fiji GPUs and way back to HD7970 and R9 280X crash the same way on Intel i7-4770K and 4790K CPUs.
I have seen things like this reported before - that it is other software in conflict.
It might be in some cases.
But I still see the crashing even if I kill all other background and foreground processes apart from game client, game, AMD Radeon Settings and other absolutely essential Windows processes.

Why is it that an Nvidia GPU can cope without having to do that on the same machine?
Not an AMD problem ....

[gmergulhao]

I have this behaviour on my setup with fresh clean windows (1909, 2004, not sure if 1803). No extra monitoring tools at all. Deepsleep bug, blackscreen before POST, BIOS time 20sec +

[Anonymous]

Problems like system instability and black screens often occur with incorrect bios settings for things like memory XMP or for things like OVERCLOCKING. which i'm absolutely certain you're doing or else you wouldnt install overclocking software like afterburner. 

The fact of the matter is, many people get a new x570 board or a newer board and grab and older gen ryzen CPU that will work with it, but they then try and set the board to use newer features like PCI express 4.0.. when the ryzen 2000 or 1000 series CPU doesnt support it on the CPU bus. The board does but the CPU doesnt, the memory controller and system communication is done by the CPU. 

Other issues are people are doing silly things like buying high end graphics cards and trying to "save power" on them.. lowering voltages also cause instability. If you want quieter and less power enable AMD chill, or use "boost" or both or something. The cards and CPU's already step up and down boost clocks or base clocks to save on power so setting power management link states and other garbage depriving your card of the power it needs to function or using overclocking software to limit fan curves and take away power will cause ... black screens and such.

Try properly enabling freesync in all 3 or 4 places needed for it to function, then try disabling enhanced sync and disabling vsync. Also ensure you've disabled every AMD graphics setting in global settings as people often turn on everything which doesnt ever work because many are dx9 or dx12 specific and others are for open GL and others are for slowing things down and limiting with the use of a buffer and others are for speeding things up and removing buffers. LEAVE IT ALL OFF IF YOU DONT KNOW.

Overlays, and overlay software and tools and importantly overclocking tools all fight for who has access and who is "first in line" and who is on the top of the stack and who is on the bottom. They will greatly slow down your system, just enabling a few graphs for performance cripples gaming FPS.. dont believe me? use doom 2016 or doom eternal VULKAN gameplay and your system in low latency freesync vsync disabled and all adrenaline global settings disabled..

then set performance monitoring to nightmare and you will see the FPS drop when more graphs and readings are added than from when its just to ultra or normal. my 2700x and x570 mainboard is quite laggy and gets 2ms-5or6 latency on average maybe? You can try disable windows 10 game mode or enable VRR in windows 10 and ensure you're up to date on windows updates and get latest AMD drivers it fixes a lot of issues with certain game titles.

Some games use their own buffering system or something so alt tabbing from them or exiting them may cause black screens.. you may need to run them full screen exclusive mode or borderless fullscreen far as i can tell.

If you want to monitor temps and FPS and other stats disable steam overlay and origin overlay and remove all the third party tools and apps and ONLY use ryzen master or the adrenaline software as it gets the job done and is probably far more accurate. If adrenaline doesnt do some things it its probably something you shouldnt be doing period. You can igors lab all you want but you dont know what you're flashing or changing to registry and editing your BIOS like that on the GPU or CPU is unnecessary risk taking. If the feature was good and necessary and stable then you can ask AMD to add the feature. If you want more performance without overclocking just enable PBO or FIT or get better RAM and NVME drives and parts in your PC, to get more performance try, buying performance parts.. Its a very unusual approach I'm sure you're all new to the idea.. like samsung branded ones or samsung bdie RAM or others listed and recommended on the AMD website. Then just fiddle with your windows 10 and its software settings or apps instead as you can easily reinstall or restore a backup or something to fix anything wrong with your OS.. fixing your PC hardware though probably isnt going to happen. At least wait till your stuff is out of warranty before you go fiddling with it trying to intentionally break it in a weeks time for about 10% faster for a week. There maybe were some issues with the "sleep" suspend on some laptops and certain systems and was often fixed by installing AMD chipset drivers like my board is x570 so i put in the ryzen x570 drivers and it comes with a ryzen power plan.. so yeah AMD power settings for your AMD system make it better a lot of the time. There may still be some bugs or issues with ryzen power plan over your system vendors bios or settings as the bios may be strangely incorrect and incompatible by default.. just try to keep your boards bios up to date and with AGESA bios updates actually bringing speed and performance boosts by a few percent you'd be stupid not to.

[triangle4]

I think what others are trying to point out is that just because it worked for you, does not mean it is the solution.

Monitoring software does much more than just accessing temp sensors.

They perform various system probes, that many other no performance related software also does.

In my case, I am a developer, and I do not have any hardware monitoring products installed, not even ryzen software.

And in the past few weeks it has happened to me 2 times.

The only part that has the power to switch off only the screen is the graphics card. Which is why most places tell you to update your graphics drivers. And most find THAT is the solution.

[Cmdr-ZiN]

@triangle4 that's a bit of a thread resurrection. 

Just search for WHEA logger error 18, Event ID error 6008 and Ryzen instability issues etc.

It's an issue affecting many many people, without an official statement people haven't been able to identify a cause.

AMD will advise to RMA your CPU after going through their troubleshooting first. Many have found a new CPU fixes the issue but many also get another CPU that doesn't work.

I'm on my 3rd CPU, the first just had instability it got better with BIOS updates but still crashed about once a week.
My second was perfect I couldn't reproduce the issue no matter how I tried but months later it became unstable and a week later it died.
My third crashed once in the first 24hrs, I updated my chipset drivers and I haven't crashed since but I'm not going to rule the problem solved, it could come back in a few months for all I know.

There's definitely an issue though, have a look for some more recent threads. Not really many answers though.

[Joeyboy]

I, for one, am grateful someone else posted this.  I have a 1 yr old AMD build exhibiting this behavior, but I don't run any of the hardware monitoring software mentioned.  I have yet to figure it out.  Nothing in the system event log, as you say.  I suspected something overheated so I popped the case cover off to ensure all fans run.  Then I downloaded speedfan to monitor temperatures and no anomalies.  I ran a cpu load test and cranked it at 100% and it ran fine.  Sometimes it fails while I am using it, just browsing websites.  Sometimes it fails when I walk away for a few hours and come back and find it in the state with no video, cpu fan running default speed.

If I figure it out I will report back.  I switched out the graphics card to a gt240, switched monitor, switched power cable, removed each DIMM.  Next I intend to try different power supply.  Sometimes it fails fast.  Sometimes it works fine for hours.  It ran well for a year before the issue emerged.  No hardware or software changes other than Windows updates.

My build

Gigabyte ds3h v2 b450m

Ryzen 3700x (not overclocked, stock hsf)

2x8gb ram (crucial ballistix)

Evga geforce 750ti graphics card