PC Graphics

[Toast21]

Ever since I upgraded my gpu from 2080super to 6900xt I get random reboots when gaming. I can be fine for 3 hours and it will reboot the system without warning. At first I thought it was the psu (750w Corsair) but I upgraded that to a 1000w Corsair psu and same problem. Gpu temps stay below 85c and cpu stays below 55c. When I had my 2080 I never experienced such problems. I deleted the nvidia drivers before installing the 6900xt too.

msi z490 gaming edge wifi 

i7 10700k

corsair rm1000 psu

gigabyte 6900xt

32gb Corsair vengeance 

Any help? I don’t know what else to do to trouble shoot this, I get no warning when it reboots, when it is running it runs great. 

[marko021994]

Hello! I have the same problem as you, have you solved it by chance?

[Tazryder]

I have a solution I have posted a detailed reply on a different site that you can probably get from my other posts but I dont wanna get flagged for spam so I removed it

but I will try and reexplain it in a textual format here, but its dependent on a particular error, check your eventlogs and look for any Dcom errors,

*** Verification Step - You Can Skip this, but it's better to make sure****

If you find them, clear the log and try and replicate the crash. Then check the log for another similar Dcom issue.

*** Verification Step - You Can Skip this, but it's better to make sure****

On Windows it's EventLog ---> Windows Logs ---> System 

Go to start ---> type "Component Services"

In Component Services in the left hand panel showing the Console Root tree

Click on the plus sign to expand the branch that says  "Component Services"

and click on the next plus sign beneath that says "Computers"

In the Middle panel/pane right click on "My Computer" to open up the Context Menu and select properties 

In the "My Computers Properties" Window, select the tab "Default Properties"

Unselect Enable Com Internet Services on this Computer

Unselect Enable Distributed Com Services on this computer

 Click OK and  restart

PS: This does NOT disable your internet, DCOM and COM is part of an old archaic system, that is rarely used in a home computer environment.

If your in a server/client environment, check with your system administrator first to see if disabling this will affect any services or legacy services that you might connect to.

[mackbolan777]

You cannot disable DCOM in Windows 10 without altering the registry and I highly advise against it. DCOM is now used for IPC processes  and services, not the old DCOM/COM you're thinking of. In fact, start trying to disable DCOM, HTTP Auto Proxy and watch your system delete your user account and you lose the system quickly. Windows takes it as an "attack" and aggressively defends by using it's internal Admin accounts to lock you out from causing harm to the point it will effectively destroy itself trying to take ownership of the files and Registry keys. If you happen to succeed in just turning DCOM and HTTP Auto Proxy off, you handicap many required Windows services. 

[Tazryder]

You can disable DCOM in windows 10..

Your wrong sir, I've had it disabled for a long time now and I'm running fine, Computers are actually quite smart now.

please check my response in detail, in no place did I state to enter the registry and I did mention to check whether you have an application that requires it. I use my PC for, development, gaming, office applications.

What on earth that's a strong aggressive reply. If I was a newb I'd be terrified to disable that now.

Where did your get that info from? " start trying to disable DCOM, HTTP Auto Proxy and watch your system delete your user account and you lose the system quickly. Windows takes it as an "attack" and " now if you perhaps own a data forensics company and need uber extreme logs or authentications or in a Windows Server /Client environment or perhaps need all of what you mention.

But I need to run my adobe products, like Photoshop and After Effects, I need to open my office products, like Word, I need to open development applications such as android studio and visual studio.

I don't need to connect to an active directory or windows AzureAD or perform any tasks that require Distributed com like

Happen to succeed? its a check box lol not an intensive system modification.  If your in dire need of it turn it back on.

A bit like a light switch sir, if it's too scary with it off, turn it back on.

[mackbolan777]

I've been selling and repairing PC's for 20 some years and passed Microsoft's software/hardware certification way back. You need DCOM. Perhaps you got lucky disabling it, you think it runs "fine". I guarantee you it does not and that my unaltered Windows would run circles around yours. DCOM has many useless features for home users and therefore tosses tons of meaningless errors, but it runs "servers" or "services" that are part of the OS that are used. You do what you want or think is "ok", it's not, ask anyone in the PC business or better yet email Microsoft support.

Besides, only Pro has the options you refer to and if you succeeded, you made a mistake. Here's why most might want it:

Do I really need Distributed Com and Component Services Administrator - Microsoft Community

[mackbolan777]

More relevant: How to disable DCOM support in Windows (microsoft.com) and why not to do it. Most of us shouldn't.

[Tazryder]

You've basically reinforced my suggestion with Windows posts, which is cool but

The windows statement also re-enforces my initial comment by mentioning this

Warning If you disable DCOM, may you may lose operating system functionality. After you disable support for DCOM, the following may result:

• Any COM objects that can be started remotely may not function correctly.

• The local COM+ snap-in will not be able to connect to remote servers to enumerate their COM+ catalogue.

• Certificate auto-enrollment may not function correctly.

• Windows Management Instrumentation (WMI) queries against remote servers may not function correctly.

These features are REQUIRED at times by admins to provide diagnostic in a client/server environment hence I said "check with your administrator" if in that environment

Local Com+ snapins would mean most of the MMC aka Microsoft Management Consoles

Certificate autoenrollment is also a Windows Server Feature typically found in places that are connected to a Windows Server, hence client and server environment

WMIs typically used in a client and server, people running things like PRTG needing up to the minute real-time diagnostics.

Your reference to IPC processes and services https://docs.microsoft.com/en-us/windows/uwp/communication/interprocess-communication in this link makes references to Pipes and COM and so forth, Named Pipes or DB aka access to databases. If you cannot use the PIPE you can always access the DB via IP.

If your a home user, I cannot see why you are unable to disable DCOM.

Again COM objects, if your in a Server Environment but if you are running Adobe Photoshop, After Effects, Office 365, Android Studio, Visual, Studio maybe remote desktop or MSTSC into another machine. No issue. ah wait if you're developing a Com+ 3.1 application you may wish to buy an NVIDIA card

Anyways sorry if I offended your domain, I thought it was a simple gamers home issue, not an IT Engineers issue, on a client connected to a Wintel or AD Environment. sorry!

[mackbolan777]

There's no practical reason to disable DCOM, unless you're annoyed by the errors it tosses. Turning it off stops a lot of things from working the way they should, things like printer sharing, devices being enumerated and stored in the HKLM area so every time you plug something into a USB port, it just works. Microsoft changed the way DCOM works a long time ago. It's built into Windows 10 Home and no, one has no access to turn it off in that version without going the regedit path, that if you don't know what you're doing, it gets messy quickly. Half the stuff you need to take ownership of belongs to "SYSTEM"(you) and "Trusted Installer"
(bot), "Owner"(machine/Microsoft), so "newbs" as you call them will ruin the OS for nothing. DCOM has no link to crashes with GPU's or really anything.

Do you not think anyone here would not have said to turn that off or Windows themselves would ship the Home version without it, if it were an issue? Some of the Com+ snap-in's com in handy and if you installed the OS all the way and turn off that IPC path, you no longer can access some security related features on Pro. Remote snap-in also does not mean "remote" as in offsite. It can mean in a different part of the drive. You do know Windows wraps around like a belt and makes 3 hidden partitions, you can't access that uses DCOM and HTTP to communicate securely? The pathways are "\\\\.\\" type format and they're entirely for IPC, System related communication. Yes the "system" talks to the other parts of the "system" using this pathway and a number  of "SID"s. 

I used to think like you, just rip it out, it wasn't like that in XP. Windows Vista and 7 weren't even totally dependent on the DCOM/HTTP Auto Proxy but alas, after several convo's with those currently developing this version of 20H2, have educated me that none of that matters and to leave it alone. Windows is based off Linux and the Unix Socket system and it does run Linux, Python, even Github type apps beneath the OS we all see. That's why the HKLM BCD is all 0's, there's leftover server 2008, Vista, even XP in the registry because the OS still uses that stuff. We just don't see it. Look at the number of SID's, none are real people but the one ending in "1001" and even that is stored in "App Data Local Temp". 

You can turn off nearly all services under "Services" and accomplish the same result, certain ones like DCOM,HTTP Auto Proxy, and a few others like anything ending in "user_numbers" cannot be turned off from there. However, you could turn off enough to cripple your PC, so if the person isn't sure about a service, it's best to make it "manual" not "disabled". 

So if anyone wants to go down the path of doing what this OP says and disable your DCOM, do it before you add a single program or driver. Do it to an installed system and you're asking for troubles.

Want to keep hackers out? Turn off "Enable Remote Connections" and "Remote Desktop". You can even disable all the "Hyper-V" stuff if you're not running any VMware. Windows 10 X64 20H2 is not designed to operate without DCOM and the rest of it's "core" components. Windows is now all called "Core" for a reason. It's so in the future things just work with Windows seamlessly. 

Disabling DCOM does nothing to prevent crashes nor does it utilize resources in excess or that you'd even notice until it's something you need. Like Xbox Gamebar or Windows Store or Tiles, Toasts, Your Phone, anything else one might want to use. COM=Common Object, so go ahead gut your OS. Has nothing to do with Azure, or any of that garbage at the end of your WOT. It's not a "simple home gamer's issue" because DCOM has zip to do with interfering with games since about XP. 

[alexv305]

I've had sam off since my last post. Not a single random reboot. Can anyone confirm this is the cause?

[mackbolan777]

I can confirm it is a possible cause since this 2.1.0.0 seems to be a problematic BIOS update. I'm going back to the 1.0.8.0 to see if that's it. The "Curve optimizer" seems to be a problem causing cores to drop/fail to boost under heavy load, even when disabled. My fight is with the 5600X and no connected issue with the RX 6800 that I could find. Seems all CPU at this point, so SAM may well be an issue. 

I have an AsRock board so this "resize bar(SAM)" option is part of the new 2.1.0.0 BIOS I did to install the 5600X. At first I had "Above 4G Decoding" on and that allowed for SAM but you lose the DVD drive due to disabling CSM. I'm sure it's an AsRock issue because their BIOS's are really twisted garbage. Ask Buildzoid. I bought mine because it was on sale. I only use Gigabyte or sometimes Asus, stuck with it for now.

[Danoisyone]

That other guy is giving you tons of flack, but ASROCK doesn't have a new BIOS for x470 and after trying 1000 different things disabling DCOM FINALLY gave me a stable round of gaming on my 6900xt.  DCOM is off and it has caused 0 issues for me, even while running all of my day to day office work . THANK YOU

[Tazryder]

WhooHoo, my DCOM solution is being recognised.. Yahoooooo

Now next step... AMD fix your **bleep** poorly coded drivers..

All DCOM is basically, is an application firewall. The AMD drivers have been wrongfully coded to request higher services it doesn't have permissions for, hence the system is shutting it down. 

People shouldn't have to disable DCOM to have a working and use able systems. 

Are these forums even checked by AMD?? 

Its just an absurd problem to have in 2021.

[AMD_NEWBIE88]

I don't usually take advice without researching the owner that posted it.

I went to google to type in Taz Ryder and I got to Taz Ryd and it autocompleted the search

https://www.crunchbase.com/person/taz-ryder  found a crunchbase with some pretty good feedback.

Graduated with a cyber security and forensics degree at 17 - WOW - genius much....are you single?

Headhunted for Facebook and GHQ, I went to Linkedin and saw contacts for both employers on your Linkedin

I can see you also own https://ackerworx.com a cyber security company, with a post about a DCOM exploit that hasn't been fixed since DCOM was invented.

So I then Googled "

mackbolan777"

 - couldn't find anything.

So I bit the bullet and disabled DCOM and I CAN USE MY PC!!!!!!!!

However, I am never buying an AMD card again. Ever!!!!!!!

[Reicoon]

Hey,

i have the same Problem, but my PC doenst reboot he shots down for a second, but the fans are still running. i need to press the power button and turn him on again. i tried everything but nothing worked yet.

did anyone found a solution?

edit: sorry for double post something went wrong

[Reicoon]

Hey,

i have the same Problem, but my PC doenst reboot he shots down for a second, but the fans are still running. i need to press the power button and turn him on again. i tried everything but nothing worked yet.

did anyone found a solution?

edit: sorry for double post something went wrong

[MOR662]

SAME PROBLEM HERE!

I'm starting to regret the moment I moved from Nvidia
To the AMD card
I had a 2080TI and everything was perfect - overall I was looking to go up a bit in FPS and game performance and since I switched to AMD 6900XT from the first day until now there is no day that was not a problem .... just problems problems problems -
Nothing like this has ever happened to me in my life - all the tickets I had in my life
Just AMD bull**bleep** and settings and need to turn it on and need to turn it off and need to check that it's in sync with it and need to turn on this setting and turn off the other - you killed my soul with all these settings - sell the card and buy 3080 the story is over
The extra performance that this card has on Nvidia is not worth the headaches and all the problems that happen with it ..... Take to the intersection of your heart friends before you buy AMD
Those who are used to Nvidia - will have a difficult life with an AMD card

[Jokerface]

Thanks so much for posting the steps regarding the DCOM.

It solved my problem and since 6 days i did not have a single crash. 

I think most people overlook this post...

I will refer to this in my original post where i stated my problem.

Although this solved my problem, i will in the future also stay clear of AMD GPU's

[SnglSrvngFrnd]

WHY DOES THIS WORK!?!? THANK YOU!!! How did you figure this out?

[sfb0519]

Hello, I have been fighting this problem for quite a while. In Cyberpunk and COD Cold War. On multiple motherboards but seems to be less frequent on older drivers from June. So what I got fed up and bought an IPMI server asrack server board. and what I just discovered since I can logon with the IPMI when it crashes and reboots the Asus 6900xt disappears from device manager and the computer has to be completely shutdown get the gpu to work again. I also experienced a crash changing to HDR in windows earlier in the night and had to shutdown remotely as the gaming pc is in the basement. I thinking of doing an RMA on the card but this seems to be a vary common issue I wish I would of bought a warrant at Microcenter that way I could of swapped on for a 3080ti or a 3090. After spending 1400 on a GPU I would gladly pay 2000+ for a card that actually works.

Asrock Rack x570d4u

AMD Ryzen 5800x no OC

16GB Gskill ram no OC

ASUS Strix 6900xt

850 Thermeltake power supply

[Anonymous]

The exchange version would have been better, but I didn't Googled for problems with my RX5700XT either. Now we have the salad. Otherwise there is a procedure that can minimize problems like these: deactivate network card, Windows Update and device installation settings, download the DDD and the Radeon driver 21.10.1 beforehand, delete all graphics drivers in safe mode with DDD and restart, monitor and Radeon graphics drivers without adrenaline software install and restart, activate network card. I don't know whether this will be successful with your particular network. I use the MSI Afterburner to set the core clock, core voltage and fan management and the FurMark graphics benchmark for testing.

[sfb0519]

Here is an update I was able to fix the problem by switching to a RTX 3080 TI. I was vary lucky I took it back to Microcenter and bagged them to let me swap it out for the 3080 and let me pay the difference they weren't going to because it was way past the 30 days because I bought it in May and I didn't purchase the warranty. But I was successful and bought a 3 year as it could be a insurance for the next generation if anyone gets my drift. But to anyone buying amd cards don't buy these you and not buy the the Microcenter warranty at least for amd cards you'll regret it if you don't and don't screw someone else by selling them a card with a potential design flaw as amd needs to address this issue as its not something we're doing to these cards.

[zebbah]

hey what fix worked for you?

[Rumbeard]

Has anyone figured this out cuz im about to RMA my gpu and i dont want it to be in vain

[mackbolan777]

I have the RX 6800 and no issue so far. However, the basics apply. Did you use DDU to remove the Nvidia and now the AMD drivers in safe mode? Have you looked in "Event Viewer" to see what crashed? Error code, etc. Start there and post back.

[Toast21]

What exactly should I post from event viewer? I’m unfamiliar with that or how to use it.

I erased all drivers earlier with ddu and the amd tool and reinstalled, and still have the same problem. I only play cod modern warfare right now, I don’t have any other games to test it on. it’s really weird because I can play for like 3 hours straight no problem and out of nowhere this happens!! I’m thinking of reinstalling windows but I don’t know if that’s even a possible solution at this point 

[delevero]

I have the excact same problem with computer that during gameplay turn it self off without warning..
And its not a heating issue and i also ran stability tests, bechmark test ect but they do not crash the computer so my system is able to handle max load and so on.. but though it turn off randomly in all games.    my gpu 6900xt and cpu 5900x cpu .

one thing that help a little was to disconnect my vr headset from the computer ( pull out all cables ) then i could play for longer before the crash.

I have now reverted to a bios named F30 from F31 hoping that the computer maybe be more stable now.

I read on the forum that people are talking about a new bios update on the way that should fix some kind of hybrid sleep mode s3 so maybe they have a fix.  https://www.techpowerup.com/276833/amd-announces-agesa-1-1-9-0-firmware-updates-improve-fclk-oc-stab...
So i think check if there are bios update and apply it when avaliable that is properbly the only thing that will fix this problem
( and maybe your video card drives as well ).
My guess is that the problem we experience could be releated to some kind of sleep mode that have a bug and trigger a shutdown

Unfortunatly a Bios update is not avaliable for my motherboard yet
https://www.gigabyte.com/Motherboard/X570-AORUS-PRO-rev-10/support#support-dl-driver

[mackbolan777]

@Coffeecat  has an idea too. Test the RAM. Not the first time corsair had a botched stick. Run Memtest X64 or swap sticks out until you find the bad ones.

[delevero]

I had some success using an older bios update, and in windows 10 power management plan i selected standard/normal and entered into advanced mode and disabled some hybrid power saving.
I was able to play for 2 houres without a  crash.
I dont know if this solved the problem or if i was just lucky.

[Toast21]

-

System

- Provider       [ Name] Microsoft-Windows-Kernel-Power       [ Guid] {331c3b3a-2005-44c2-ac5e-77220c37d6b4}

EventID 41

Version 8

Level 1

Task 63

Opcode 0

Keywords 0x8000400000000002

- TimeCreated       [ SystemTime] 2021-01-16T06:56:34.1618370Z

EventRecordID 25143

Correlation

- Execution       [ ProcessID] 4       [ ThreadID] 8

Channel System

Computer DESKTOP-5T752EE

- Security       [ UserID] S-1-5-18

-

EventData

BugcheckCode

270

BugcheckParameter1

0x36

BugcheckParameter2

0xffff8f85e85f9810

BugcheckParameter3

0xffffa201b44137e8

BugcheckParameter4

0xffffa201a9773f40

SleepInProgress

0

PowerButtonTimestamp

0

BootAppStatus

0

Checkpoint

0

ConnectedStandbyInProgress

false

SystemSleepTransitionsToOn

0

CsEntryScenarioInstanceId

0

BugcheckInfoFromEFI

false

CheckpointStatus

0

CsEntryScenarioInstanceIdV2

0

LongPowerButtonPressDetected

false

does that give you any valuable info? 8 seconds after that this error shows as well saying pc rebooted from a bug check

-

System

- Provider       [ Name] Microsoft-Windows-WER-SystemErrorReporting       [ Guid] {ABCE23E7-DE45-4366-8631-84FA6C525952}       [ EventSourceName] BugCheck

- EventID 1001       [ Qualifiers] 16384

Version 0

Level 2

Task 0

Opcode 0

Keywords 0x80000000000000

- TimeCreated       [ SystemTime] 2021-01-16T06:56:42.8680238Z

EventRecordID 25187

Correlation

- Execution       [ ProcessID] 0       [ ThreadID] 0

Channel System

Computer DESKTOP-5T752EE

Security

-

EventData

param1

0x0000010e (0x0000000000000036, 0xffff8f85e85f9810, 0xffffa201b44137e8, 0xffffa201a9773f40)

param2

C:\Windows\MEMORY.DMP

param3

fc807c2f-5c51-412b-8328-0f7eec79d18b

[mackbolan777]

@Toast21 

That shows that bug check 270 and the related parameters are the video driver attempting to remove a device that is still connected, in other words, recover from failure. Bug Check 1001 is common after a hardware failure of any kind and references a page not found in memory. Basically this doesn't say a lot. I like the layout you did, very nice.

Ok, so go into Radeon software and select "Performance", "Tuning", "Automatic". Accept the warning, it's a non-binding disclaimer. Select "undervolt GPU" and write down that number. Select "manual" in the same area. Enable all the disabled tuning things. Slide the minimum GPU clock so it's within 100Mhz of the maximum. Enter the voltage you wrote down into the voltage area or slide the bar to match that number. Disable "zero RPM" fan mode. Set the "power limit" to maximum. Select "Apply" at the top right. Click the 3 dots at the top right to save the profile in the default location. Also disable any CPU OC if you're running one temporarily for this test. If your ok, put the CPU OC back on and retest.

What this does is it stop the GPU from cycling the clock up and down so drastically which prevents voltage spikes, improves the voltage for cooler temps, makes the fans spin full time at a low speed to reduce idle temps. The power limit is only used as needed by the card under load, it's like a reserve.

Should that not work, download Memtest from:  MemTest: RAM reliability tester for Windows (hcidesign.com)

You can use the free version or buy the deluxe for $14 which allows you to make a bootable USB/DVD/CD to fully test the RAM and CPU cache's outside of Windows running. I personally bought it and found it very useful. You can then run tests that go past 100% coverage to find any intermittent failures. Memtest x86 is another that works ok. Both have free or paid versions. 

@Coffeecat  Had mentioned a link about someone having some bad RAM. It's not too far out of an idea that a stick is bad or that running 32GB along with 16GB of VRAM is causing a problem, where a CMOS reset might fix that or going down to 16GB might fix it too. That's more of an AMD Ryzen issue though. 

Lastly, the PSU might be tripping it's internal protection for voltage spikes or drops related to the powerful card. This will cause the shutdown/reboot as well. If that's the final case, try a different PSU. Corsair RMx in the wattage you need is a good choice. Having a PSU older than 10 years can be an issue. But let's see if the above work around fixes it or you find bad RAM first. You can reference my screen shot of my settings, a page back, to see how I did mine, your settings will be different. 

[Verk]

Has anybody found a fix for this. I'm having the exact same issue. My specs are a Ryzen 5950X, An Asus TUF gaming plus Motherboard, 32 GB of 3200Mhz DDR4, 1000w Corsair RMX PSU and a 6900XT.

The problem for me started when I upgraded my monitor from a 1080p to a 1440p one.

Sometimes the system will reboot after gaming for 5 minutes other times it's fine for close to an hour.

At first I thought it was a faulty graphics card and the company I got the PC from swapped it but the problems persisted.

Then they had the PC in for repair and swapped the PSU and motherboard but it's still doing it.

[marcelokayro]

One more with the same problem.

Reboots, some time my GPU usage goes to 80%/90% without any 3D application(I already disable all transparency and etc).

It's annoying. ,

Setup: AMD 5900X + 6900XT Phantom OC  + ASROCK Taichi.

Before, I started with this issue when I replace my old Asus Tuf X570 to the ASROCK. 

I did a fresh install multiple times and still happening.

[Toast21]

Update: I returned the old gigabyte 6900xt and picked up another one. Same exact issue! Even reinstalled windows. I tried everything so far that has been discussed in this thread but still crashing. In the time I returned my first 6900xt and picked up another, I threw my 2080super back and it was running flawless for a good 2-3 weeks, it makes me believe it’s a driver issue and not a hardware issue. I don’t know what to do at this point other than looking for a 3080/3090 if this problem persists. I really want to use this card I just can’t stand these system crashes!

[mackbolan777]

Have you used DDU to remove the Nvidia drivers and rerun it to remove the AMD drivers in safe mode?

Display Driver Uninstaller Download version 18.0.3.6 (guru3d.com)

You might need to run it twice to remove both drivers. All else fails, reinstall the OS with the latest Version 20H2. Use a blank 8GB USB drive and the Windows Media Creation Tool, then pick you version during setup. Choose to make media for another PC or device instead of for YOUR PC. This will give you a full, clean updated version without pulling files off of your PC that may not be up to date or corrupt.

Of course use 2 X Pcie power cables. Make sure the GPU is fully seated, especially in the front where the metal tabs need to fully drop into the case. I'm using the same driver, albeit an RX 6800, the same driver and it's a Gigabyte as well. I think you have some OS related software problem or improperly uninstalled/conflicting driver problem. 

[jtgolden1s]

Hello, any update on this? I went from a 2080ti to a 6900xt and started having reboots when I turn on my PC. I have tried everything including reinstalling windows. I don't want to RMA the card because I don't believe that is the issue with it. I think it is software or something going on in my PC.

[mackbolan777]

Full system specs please.

[Gexx]

Is your 6900XT a 2x8pin or 3x8pin (ASROCK 6900XT Phantom gaming D) like mine?

Im having weird issues as well, but nowhere near the degree you're at.

[alexv305]

Hello, I'm having the same issue as well. 5800x, x570 tomahawk, 6900xt. Memtest my 32gb of ram, runs flawless. 

Games will run fine for 1-2 hours then system will reboot with no warning. Also happens at random, during idle, or light work loads as well. No windows crash or anything just a reboot. Tried a different power supply thinking it was power spikes, still same issue. Only thing that seems to work is disabling Smart Access Memory. So far a couple days without a crash with SAM disabled. 

Anyone else on here can confirm?

[Jeru]

Hi Toast21

Do you have an news regarding the reboot problem as i have the exactly same issue.

Thanks Alex