PC won't boot after trying to fix recurrent black screens of death whilst gaming

[Jackrkl]

TLDR: Since building the PC I have had recurrent black screens of death when: gaming, running windows memory test. Tried moving the RAM sticks to see if that was the problem. PC no longer boots (case fan LEDs turn on, BUT no case or CPU fans are spinning, I can hear the PSU fan spin on)
To make the rest easier to read I have split it up based on the problems I encountered.


Specs:

PSU: 660W Fractal Design Ion+ 660P, Full Modular, 80PLUS Platinum, SLI/CrossFire, Single Rail, 55A, 140mm Fan, ATX PSU 6Z4BJ​

​

CPU: INTEL CORE I5-11400F 2.60GHZ (ROCKET LAKE) SOCKET LGA1200 PROCESSOR - BX8070811400F​

​

GPU: 3060ti FE​

​

Motherboard: MSI MAG B560M MORTAR WIFI MATX Motherboard​

​

Ram: Crucial Ballistix 3600Mhz 16GB (2x8GB) Gaming Memory Black​

​

Case: Kolink Citadel Mesh RGB Micro-ATX Case – Black (3 case fans 2 in 1 out)​

=========================================================================================================================

Background:

This PC was built by myself In July 2021. This was my first time building a computer. No problems encountered during the build. I have been unable to use the PC frequently so its barely been used. I've only had time now to set it up and try and solve this problem.

=========================================================================================================================

Booting problems:

• When I first built the PC I noticed it booted weirdly. Didn't think much of it as it would still load up. Problem described below:
  • PC would power on for 1-2 seconds. Then power off for 1-2 seconds. Then usually fully power on and launch windows
  • This behaviour would also occur during the reboot after the black screen of death (detailed below)
• Just before my PC stopped loading/turning on it would get into boot cycles where the booting for 2 seconds followed by turning off would occur 3-5 times in a row
• I was able to film this behaviour on my phone video here :

  http://imgur.com/a/LaiiAK7

• 1st video (top of the imgur page) occurs during metro exodus resulting in a black screen, after which the PC attempts to reboot and fails (like it does constantly during powering on), after which it fully reboots successfully. This was the original problem I was trying to fix
• 2nd video (bottom of imgur page) is of a boot loop/ reboot cycle which occured near the end (before I messed with the ram). The PC was constantly trying to boot, dying and restarting again. This occured shortly before the PC refused to start up
• During these reboot attempts the motherboard EZ Debug LEDs would reach the DRAM LED after which the light would turn off and the PC would restart again (can be seen in the video). After failing once it would usually pass through all of the Motherboard LEDs and boot successfully. However near the end the PC was caught in a repeated boot cycle/restart loop.
• Note at no point did I get a constant EZ debug LED light which I am aware indicated a problem. I am aware that these lights usually turn on and off during booting. I thought the pattern might help troubleshooting

Lights on my motherboard are coloured​

• CPU = red light
• DRAM = orange light
• VGA = white light
• BOOT = very tiny green light (can't be seen in the videos)

​

=========================================================================================================================

Black screen of death problem:

• Got these from as soon as I built the computer. Initially I thought this was due to my undervolting the GPU and increasing the RAM clock speed. Discovered in the event log a number of Kernel-Power Critical error 41 events
• Black screen would cause the PC to reboot and have the same booting behaviour explained above "power on for 1-2 seconds. Then power off for 1-2 seconds. Then usually fully power on and launch windows"
• The vast majority of these black screens of death have occurred during gaming. For more intense games like metro exodus it occurs immediately, for older less intense titles it may occur after some play time
• Whilst troubleshooting this black screen also occurred during Window Memory check (twice, I was never able to complete the diagnostic)
• Near the end (1/2 hours before the PC would no longer turn on/boot) I did have the black screen of death occur during: MSI BIOS and when browsing the internet. Neither had triggered a black screen in the past
• After one of the black screens I did get the following error: "Memory overclock fail! Press F2 to load default values and continue"

> I pressed F2 but I believe I got another black screen of death during this and my PC rebooted again. Not sure if these settings went in. I did however reset my changes in the MSI BIOS and turn of XMP profile​

• Because of the black screens during Windows RAM check I figured something was wrong with the RAM sticks. I decided to inspect and move the RAM sticks around in the case. Since then the PC does not boot (problem described in detail below)

=========================================================================================================================

Not booting:

• Currently the case fans LEDs do turn on. BUT the fans of the CPU and case do NOT spin
• No other lights present on the motherboard (normally during boot up the ez Debug LEDs would turn on during the boot up temporarily).
• I can hear the PSU spinning
• The case LED indicators next to the power button do not turn on (not 100% of this as I didn't pay too much attention to this before).
  • I believe the two LEDs indicate: power LED, SSD/HDD activity. Mine are shown in this image GEKL_076_05-opti.jpg (800×800) (kolink.eu) <-- they are to the right of the power button in this image

• Problem with not starting has occured since I inspected the RAM, tried 1 RAM stick and moved the sticks to see if the RAM was the problem
• Initially it sounded like the PC was booting (CPU and case fans were on) but I couldn't get it to connect to my monitor. Now it no longer boots at all (just case fan LEDs turn on)

• I tried resetting the CMOS twice, checking everything was plugged in correctly to the PSU and motherboard
• Started the PC by shorting the motherboard pins to confirm the case power button was working

=========================================================================================================================

What I've tried:

• Windows memory test - tried to run twice but got a black screen during both tests
• Intel’s Processor Diagnostic tool - downloaded but my computer would not launch the application despite launching as administrator
• Updated all drivers (including latest Nvidia drivers)
• Ran command prompt “sfc /scannow” which did not find any integrity violations
• Reset/cleared the BIOS
• Reset the CMOS
• Moved the ram and tried 1 ram stick
• Used a friends RAM stick
• Shorted the power button on the motherboard to see if it was the case
• Temperatures seem fine I have included screenshots of hardware monitor and MSI afterburner link here:

  http://imgur.com/a/LaiiAK7

Any help would be greatly appreciated! As I am a bit lost what to do

 

[eidairaman1]

Try a different power supply for starters

 

[Count von Schwalbe]

• Used a friends RAM stick

What speed/type of RAM?

Have you tried booting with JDEC-only values (AKA turn off XMP)?

When it boots, does HWiNFO show voltage values that are too low or high?

 

[qubit]

Try a different power supply for starters

What this one says. ^^

When that doesn't fix it, do the following:
- Remove all non essential components, including the graphics card, using the IGP instead. Remove all drives, cards, USB devices except keyboard and mouse, etc. We're trying to achieve the baseline of a stable boot first
- Run off ONE stick of RAM
- Load BIOS defaults if it boots ok and check if an update is available, but don't flash it

Let us know how you get on and we'll move to the next step.

 

[Jackrkl]

What speed/type of RAM?

Have you tried booting with JDEC-only values (AKA turn off XMP)?

When it boots, does HWiNFO show voltage values that are too low or high?

Ram is 3600Mhz

I had set the XMP profile to 3600Mhz when I first built the PC. I turned off XMP but the PC stopped working shortly after so wasn't able to test if this made a difference.
I also got a Memory overclock fail! Press F2 to load default values and continue. I tried pressing F2 but I think I got another black screen of death so the change may not have gone through

Unfortunately I didn't check HWiNFO before the PC crashed. Since it won't turn on/boot at the moment I will check this as soon as I find a solution to the booting problem

Thank you for the help though, I hadn't realised sometimes the motherboard/CPU impacts the speed you can run your ram. I was following guides which said to set the XMP profile to your rams speed

Try a different power supply for starters

Will do. This was my first build so don't have a PSU easily accessible. Would you recommend returning the one I currently have (not sure if it'll be under warranty etc). Also do you think 660 watts would be sufficient for this build as I was worried this maybe a limiting factor

What this one says. ^^

When that doesn't fix it, do the following:
- Remove all non essential components, including the graphics card, using the IGP instead. Remove all drives, cards, USB devices except keyboard and mouse, etc. We're trying to achieve the baseline of a stable boot first
- Run off ONE stick of RAM
- Load BIOS defaults if it boots ok and check if an update is available, but don't flash it

Let us know how you get on and we'll move to the next step.

Thank you so much. I will let you know once I'm able to source another PSU as I don't have any at home. Do you think if I went to a repair shop they'd allow me to use one of theirs? or should I buy a new PSU and hope it works. Money is a bit tight atm so want to minimise spending.
I've asked some friends but they only have 500 watt PSUs

 

[Hofnaerrchen]

This was my first build so don't have a PSU easily accessible. Would you recommend returning the one I currently have (not sure if it'll be under warranty etc). Also do you think 660 watts would be sufficient for this build as I was worried this maybe a limiting factor

Try to get your friend's PSU.

 

[Count von Schwalbe]

Since it won't turn on/boot at the moment

Sorry, I missed that bit.

Thank you for the help though, I hadn't realised sometimes the motherboard/CPU impacts the speed you can run your ram. I was following guides which said to set the XMP profile to your rams speed

Yes, these are very important parts of RAM OC. 3600 isn't too far out there though, so it should have been ok.

With Ryzen, you can push a little more voltage to the IMC, but IDK if you can do that with Rocket Lake.

I will let you know once I'm able to source another PSU as I don't have any at home. Do you think if I went to a repair shop they'd allow me to use one of theirs? or should I buy a new PSU and hope it works. Money is a bit tight atm so want to minimise spending.
I've asked some friends but they only have 500 watt PSUs

If you don't overload the GPU or CPU too much, a 500 watt should work for testing. Make sure to confirm voltages on the new one if it boots, we can help compare them to where they should be, in case the voltage regulation is off on the motherboard.

 

[qubit]

Thank you so much. I will let you know once I'm able to source another PSU as I don't have any at home. Do you think if I went to a repair shop they'd allow me to use one of theirs? or should I buy a new PSU and hope it works. Money is a bit tight atm so want to minimise spending.
I've asked some friends but they only have 500 watt PSUs

Yer welcome.

That 500W PSU will be fine, especially without the graphics card, so please do try that first. It should power that card just fine anyway, especially in 2D mode, ie the desktop.

Also, to clarify, if it doesn't boot in that minimum configuration, try another RAM stick in another slot.

 

[Count von Schwalbe]

without the graphics card

With an 11400F that could be fun.

 

[Jackrkl]

If you don't overload the GPU or CPU too much, a 500 watt should work for testing. Make sure to confirm voltages on the new one if it boots, we can help compare them to where they should be, in case the voltage regulation is off on the motherboard.

Thank you I think I'll ask my friend for his PSU to try then. My theory was the black screen was being caused under load since it mainly occurred during gaming. Does this mean I wouldn't be able to rule this out as I wouldn't be able to put the 500 watt PSU under much strain. Therefore I won't be able to see if the change in PSU resolves the black screens?

That 500W PSU will be fine, especially without the graphics card, so please do try that first. It should power that card just fine anyway, especially in 2D mode, ie the desktop.

Also, to clarify, if it doesn't boot in that minimum configuration, try another RAM stick in another slot.

Perfect I'll try the RAM trick if that doesn't work. Would I be able to boot the minimum system without the GPU? As my CPU doesn't have any integrated graphics. I know you can eventually boot systems without any ram sticks but I'm not sure if the CPU could cover the lack of a GPU

Try to get your friend's PSU.

Will do hopefully 500 watts doesn't make it worse. Can you damage components by using too low a wattage CPU?

 

[Count von Schwalbe]

Thank you I think I'll ask my friend for his PSU to try then. My theory was the black screen was being caused under load since it mainly occurred during gaming. Does this mean I wouldn't be able to rule this out as I wouldn't be able to put the 500 watt PSU under much strain. Therefore I won't be able to see if the change in PSU resolves the black screens?

If the issue was happening at boot, load shouldn't matter. Let's get you PC booted and see if we can identify any other issues.

Would I be able to boot the minimum system without the GPU?

No iGPU, you need dGPU.

Can you damage components by using too low a wattage CPU?

I assume you meant PSU, and the answer is no. Just try not to pull too much or it could shut down randomly.

Looking into your build again, you should be fine with a 500W, even with some stress testing. Just don't do furmark and Prime 95 together.

 

[Jackrkl]

If the issue was happening at boot, load shouldn't matter. Let's get you PC booted and see if we can identify any other issues.

Perfect will let you know if I can boot the PC using the alternate PSU. To clarify I should have the GPU plugged in as well whilst doing this?

Looking into your build again, you should be fine with a 500W, even with some stress testing. Just don't do furmark and Prime 95 together.

If I do manage to boot with the 500w PSU. Should I then try and replicate the black screen errors I was getting underload. To do this would you recommend furmark or Prime 95? I was using heaven whilst trying to undervolt the GPU
Also would it be sensible to put my ram underload to see if that is a problem?

Thank you so much for the help and Happy Christmas

 

[eidairaman1]

Ram is 3600Mhz

I had set the XMP profile to 3600Mhz when I first built the PC. I turned off XMP but the PC stopped working shortly after so wasn't able to test if this made a difference.
I also got a Memory overclock fail! Press F2 to load default values and continue. I tried pressing F2 but I think I got another black screen of death so the change may not have gone through

Unfortunately I didn't check HWiNFO before the PC crashed. Since it won't turn on/boot at the moment I will check this as soon as I find a solution to the booting problem

Thank you for the help though, I hadn't realised sometimes the motherboard/CPU impacts the speed you can run your ram. I was following guides which said to set the XMP profile to your rams speed


Will do. This was my first build so don't have a PSU easily accessible. Would you recommend returning the one I currently have (not sure if it'll be under warranty etc). Also do you think 660 watts would be sufficient for this build as I was worried this maybe a limiting factor


Thank you so much. I will let you know once I'm able to source another PSU as I don't have any at home. Do you think if I went to a repair shop they'd allow me to use one of theirs? or should I buy a new PSU and hope it works. Money is a bit tight atm so want to minimise spending.
I've asked some friends but they only have 500 watt PSUs

With what you just said, go to a shop

 

[izy]

Perfect will let you know if I can boot the PC using the alternate PSU. To clarify I should have the GPU plugged in as well whilst doing this?



If I do manage to boot with the 500w PSU. Should I then try and replicate the black screen errors I was getting underload. To do this would you recommend furmark or Prime 95? I was using heaven whilst trying to undervolt the GPU
Also would it be sensible to put my ram underload to see if that is a problem?

Thank you so much for the help and Happy Christmas

Try booting and testing after you resetted the bios to default and go from there with what everyone else said if the problem is not fixed. (Remove bios battery)

 

[Jackrkl]

Dear All,

Merry Christmas

Progress!

Since my last post I've managed to get the PC booting again! I believe the problem was caused by accidently loosening one of the PSU cables to the motherboard when moving RAM around.
Whilst tinkering with the components today I also removed the CMOS battery and shorted the CMOS pins again.
The PC now successfully boots and I turned the XMP profile off.

I believe the XMP profile was the cause of the problems I was having. After researching this problem further I have found out that the INTEL CORE I5-11400F 2.60GHZ does NOT support RAM at speeds above 3200MHz
When I initially built the PC I had set the XMP profile to 3600MHz (my RAM sticks advertised speed)
After I reset the CMOS the bios set the RAM speed to 2666MHz which I am now running currently. I believe this is the default DDR4 Ram speed.
I have seen some reviews and benchmarks of the i5 11400f with RAM speeds above 3200MHz. Is there any reason why this works? I understand the CPU has a memory controller which restricts the MHz the RAM cna run at (depending on the CPU). I figured going over 3200MHz would cause the instability I was experiencing

Since the CMOS reset I have been able to run applications which would normally result in a black screen of death and reboot e.g.

• Windows memory diagnostic which "detected no errors" [before this did not complete after 2 attempts due to black screen of death]
• Metro exodus enhanced edition benchmark finished when previously it would crash as soon as it started

I still occasionally get some failed boots when turning the PC on. This video is from when I was stuck in a boot loop so NOT exactly whats happening but the sequence of the EZ debug LEDs is the same when booting:

http://imgur.com/a/eS7LG84

• PC would power on for 1-2 seconds (go from Red LED and reach the orange DRAM light). Then it would power off for 1-2 seconds. Then the PC would fully power on and launch windows (goings from red to orange to white to a faint green whcih isn't detected in the video I have linked)

Since I have the PC working again I wanted to ask your advice for next steps to make sure nothing else is wrong with the system?

1. Are there any benchmarks I should run to test the system ram, cpu, PSU etc to make sure there are no other problems
2. Should I update my BIOS to a later version (I'm currently running the one from Summer 2021 but there are some later versions on the motherboard page)
3. Would it be safe to boost my RAM speed to 3200MHz? As the intel RAM limit for my CPU is apparently 3200MHz

Thank you all again for the amazing help. This experience has really taught me a lot
Thanks again!

If the issue was happening at boot, load shouldn't matter. Let's get you PC booted and see if we can identify any other issues.

No iGPU, you need dGPU.

I assume you meant PSU, and the answer is no. Just try not to pull too much or it could shut down randomly.

Looking into your build again, you should be fine with a 500W, even with some stress testing. Just don't do furmark and Prime 95 together.

Yer welcome.

That 500W PSU will be fine, especially without the graphics card, so please do try that first. It should power that card just fine anyway, especially in 2D mode, ie the desktop.

Also, to clarify, if it doesn't boot in that minimum configuration, try another RAM stick in another slot.

Try to get your friend's PSU.

Try booting and testing after you resetted the bios to default and go from there with what everyone else said if the problem is not fixed. (Remove bios battery)

With what you just said, go to a shop

 

[qubit]

Great it's working again. When you have achieved total stability, then it's wise to update the BIOS.

I wouldn't run the RAM faster than what the CPU is rated for as you're likely to see instability, if not immediately, then likely over time.

You could try Prime95 which is a good CPU stress test. Keep a close eye on temperature though and stop it if it starts to overheat.

 

[izy]

Dear All,

Merry Christmas

Progress!

Since my last post I've managed to get the PC booting again! I believe the problem was caused by accidently loosening one of the PSU cables to the motherboard when moving RAM around.
Whilst tinkering with the components today I also removed the CMOS battery and shorted the CMOS pins again.
The PC now successfully boots and I turned the XMP profile off.

I believe the XMP profile was the cause of the problems I was having. After researching this problem further I have found out that the INTEL CORE I5-11400F 2.60GHZ does NOT support RAM at speeds above 3200MHz
When I initially built the PC I had set the XMP profile to 3600MHz (my RAM sticks advertised speed)
After I reset the CMOS the bios set the RAM speed to 2666MHz which I am now running currently. I believe this is the default DDR4 Ram speed.
I have seen some reviews and benchmarks of the i5 11400f with RAM speeds above 3200MHz. Is there any reason why this works? I understand the CPU has a memory controller which restricts the MHz the RAM cna run at (depending on the CPU). I figured going over 3200MHz would cause the instability I was experiencing

Since the CMOS reset I have been able to run applications which would normally result in a black screen of death and reboot e.g.

• Windows memory diagnostic which "detected no errors" [before this did not complete after 2 attempts due to black screen of death]
• Metro exodus enhanced edition benchmark finished when previously it would crash as soon as it started

I still occasionally get some failed boots when turning the PC on. This video is from when I was stuck in a boot loop so NOT exactly whats happening but the sequence of the EZ debug LEDs is the same when booting:

http://imgur.com/a/eS7LG84

• PC would power on for 1-2 seconds (go from Red LED and reach the orange DRAM light). Then it would power off for 1-2 seconds. Then the PC would fully power on and launch windows (goings from red to orange to white to a faint green whcih isn't detected in the video I have linked)

Since I have the PC working again I wanted to ask your advice for next steps to make sure nothing else is wrong with the system?

1. Are there any benchmarks I should run to test the system ram, cpu, PSU etc to make sure there are no other problems
2. Should I update my BIOS to a later version (I'm currently running the one from Summer 2021 but there are some later versions on the motherboard page)
3. Would it be safe to boost my RAM speed to 3200MHz? As the intel RAM limit for my CPU is apparently 3200MHz

Thank you all again for the amazing help. This experience has really taught me a lot
Thanks again!

Your MB is probably re-training the RAM when it doesnt boot (maybe it just doesnt like you RAM that much and you have to increase some volts for a higher frequency). You can try using "Memory try it" from the OC menu with a decent speed and maybe that will fix it but you will have to look for the right timings and tweak them after you find a stable frequency with Memory try it because they will probably be very loose.

Edit:
Update the BIOS , that could give you some memory stability , they usually fix some RAM bugs with BIOS updates
You can try using TM5 and OCCT to check for memory stability and you can use AIDA to benchmark.
You can try "Memory try it" with with 3200MHZ , seems that your CPU should be able to run it. (try higher timings option first) Also enable memory retry or what is called , 3 - 4 time is enough before MB will try to recover automatically if its failing to boot. (seems its called Memory Failure Retry / Memory Retry Count) - usually "Memory try it!" will adjust the voltages too.
EDIT2:
If it doesnt want to work with 3200Mhz it might be a problem and you can try (just for testing) to move your sticks in different slots , normally you should have them in slot 2 and 4 counting from the CPU. You can also try with 1 stick (just to be sure the RAM slot / sticks are OK and in working condition)
I think you can have you RAM working at higher frequency if you use Gear2 but i dont think it will be faster than Gear1 at a lower speed.

 

[maxfly]

The next time you power off your rig. Disconnect the power cable from the wall. Then go through each cable in your rig, one by one, and make doubly sure they are plugged in nice and tight. Particularly your cpu fan cables.
Are you absolutely certain that all of your motherboard standoffs were used in the correct holes and screwed in completely? If not, it would be a good idea to pull your mb out of the case and check that you don't have an extra standoff or a screw stuck in between the mb and case or a standoff that's pushing your mb up.
GL!

 

[KLiKzg]

Since I have the PC working again I wanted to ask your advice for next steps to make sure nothing else is wrong with the system?

1. Are there any benchmarks I should run to test the system ram, cpu, PSU etc to make sure there are no other problems
2. Should I update my BIOS to a later version (I'm currently running the one from Summer 2021 but there are some later versions on the motherboard page)
3. Would it be safe to boost my RAM speed to 3200MHz? As the intel RAM limit for my CPU is apparently 3200MHz

Thank you all again for the amazing help. This experience has really taught me a lot
Thanks again!

So, back again to basics...your system will run based on lowest possible speed available. So the RAM setting was wrong to use 3600MHz, as the MBO & CPU can't handle it.

Let me help answer your questions:
1. Yes there are benchmarks to run. For RAM use MEMtest & do not give up until it has run all the way without errors - or you will go back to beginning of this topic.
Also, I personally use FREEware software like HIREN's boot version. Hope it will help you out.
2. Answer for that is in the LOG FILE of changes for the updated BIOS. Please read more, unless you want more BSOD (blue or black).
3. Answer to that question is to find the max MHz RAM speed for your MBO & for your CPU - lowest of that settings is the right for you (unfortunately, Life is no fair). Also, please check "ark intel web pages" for more info about CPU & MSI web pages for your MBO - download the guide for your MBO & read it.

& forget overclocking, unless you have some 10+ years of experience...or you will ruin your equipment.

Enjoy the testing.

 

[Hofnaerrchen]

I am happy for you to have a running system again. Congratulations!

The way you did present your problem was very exemplary, as well as your report on the found results.

According to memory speeds. By my own experience memory speeds/timings do not impact daily performance much after reaching a certain sweet spot which is - for most DDR4 configurations - 3200MHz CL16.

For testing memory stability the Windows memory test should do the job. But if you really want to do a hardcore test, you can run memtest86 from a bootable USB-drive. This will take quite some time so it's a good idea to let it run while you are at sleep or work.

 

[eidairaman1]

In the end there are so many variables in a users rig config that YMMV.

 

[delshay]

TL,DR Not sure if the OP has the same issue I had, but I had black screen when playing games or just sometimes watching Youtube videos. The issue in my case was the HDMI cable.
The cable itself worked fine with other card but will sometimes cut-out when using Vega card. Swapping the cable to slightly shorter higher quality cable fixed my issue, ie no more intermittent black screen cut-out.

 

[Jackrkl]

Thanks to everyone for your insightful answers and support! Will defo be using this forum again if I need any help!!!

For future readers reading this thread problem was causing by my intel CPUs (i5 11400f) having an upper RAM speed limits (3200MHz). I had bought 3600,Hz ram and had se tthe XMP profile to 3600MHz. This was causing instability resulting in black screen of death in gaming and heavy loads/browsing. Problems with failed booting and boot cycling was caused by motherboard re-training the RAM due to setting too high a speed.


=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================

Great it's working again. When you have achieved total stability, then it's wise to update the BIOS.

I wouldn't run the RAM faster than what the CPU is rated for as you're likely to see instability, if not immediately, then likely over time.

You could try Prime95 which is a good CPU stress test. Keep a close eye on temperature though and stop it if it starts to overheat.

Thank you so much. I followed your advise and updated the bios to the latest version. I have xmp profiles on but ONLY at 3200MHz.

Should I install the latest drivers for the motherboard? When I was downloading the BIOS I noticed there were some newer verisons?

Your MB is probably re-training the RAM when it doesnt boot (maybe it just doesnt like you RAM that much and you have to increase some volts for a higher frequency).

Seems to have solved the black screens and even the problems when booting the PC! Have not had the same problem where when powering on it would result in a restart. I think this boot problem was caused by the RAM speed being too fast causing the motherboard to manually reduce its speed (causing the failed boot attempt) and then restarting with a slower speed on the RAM ( like you said motherboard re-training the RAM)

PC seems very stable now as I was undervolting my GPU and doing multiple stress/benchmark tests and I had no kernel errors

You can try using "Memory try it" from the OC menu with a decent speed and maybe that will fix it but you will have to look for the right timings and tweak them after you find a stable frequency with Memory try it because they will probably be very loose.

Since I now have the RAM stable is there any point trying memory try it? The motherboard and cpu seem a lot happier now. my heart still skips a beat during booting but thankfully no problems to report

You can try "Memory try it" with with 3200MHZ , seems that your CPU should be able to run it. (try higher timings option first) Also enable memory retry or what is called , 3 - 4 time is enough before MB will try to recover automatically if its failing to boot. (seems its called Memory Failure Retry / Memory Retry Count) - usually "Memory try it!" will adjust the voltages too.
EDIT2:
If it doesnt want to work with 3200Mhz it might be a problem and you can try (just for testing) to move your sticks in different slots , normally you should have them in slot 2 and 4 counting from the CPU. You can also try with 1 stick (just to be sure the RAM slot / sticks are OK and in working condition)
I think you can have you RAM working at higher frequency if you use Gear2 but i dont think it will be faster than Gear1 at a lower speed.

Thank you so much for such a detailed reply. My Bios looked just like this although I didn't alter the RAM speed using memory try it. Instead I used the DRAM frequency option. Is this correct?


=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================

The next time you power off your rig. Disconnect the power cable from the wall. Then go through each cable in your rig, one by one, and make doubly sure they are plugged in nice and tight. Particularly your cpu fan cables.
Are you absolutely certain that all of your motherboard standoffs were used in the correct holes and screwed in completely? If not, it would be a good idea to pull your mb out of the case and check that you don't have an extra standoff or a screw stuck in between the mb and case or a standoff that's pushing your mb up.
GL!

All screwed in when I checked. Yes I think the motherboard cable being quite chunky was not connected properly. I also didn't really get a satisfying click when connecting. Im not sure if thats unusual or not. I think the loose motherboard cable was causing the failing to boot and only case fan LEDs being present as they are NOT powered via the motherboard. Instead there was a PCB fan LED controller which came with the case and plugged directly into the PSU. Thats why this was the only one working


=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================

Answer for that is in the LOG FILE of changes for the updated BIOS. Please read more, unless you want more BSOD (blue or black).
3. Answer to that question is to find the max MHz RAM speed

Thank you so much. The BIOS update was a bit nerve racking but has seemed to have helped as well as turnign down the RAM speed. No overclocking for me, I'm onyl plannign to undervolt the GPU for better temps, less power draw. Does that count as an OC? I heard that undervolts couldn't damage the card as you'd get a blue screen of death if you powered it too low


=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================

I am happy for you to have a running system again. Congratulations!

The way you did present your problem was very exemplary, as well as your report on the found results.

According to memory speeds. By my own experience memory speeds/timings do not impact daily performance much after reaching a certain sweet spot which is - for most DDR4 configurations - 3200MHz CL16.

For testing memory stability the Windows memory test should do the job. But if you really want to do a hardcore test, you can run memtest86 from a bootable USB-drive. This will take quite some time so it's a good idea to let it run while you are at sleep or work.

Thank you so much! I tried posting on reddit and the windows forum and only got the odd reply which wasn't too helpful. Everyones replies here have been very helpful.
Glad the things i posted were helpful. Hopefully they can help others in the future with similar problems.
Wish I'd known about the intel RAM limits!!!! Never heard anyone mention them could have saved a few £ but no big deal. I'm just happy I don't have an expensive repair bill
Windows memory test hasn't found any errors thankfully. I've also used superposition benchmark and metro exodus on Ultra 1440p with no crashes/black screen of death so I am happy with the stability for now


=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================

In the end there are so many variables in a users rig config that YMMV.

Exactly, this forum has been amazing at suggesting ideas and helping me narrow it down



=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================
=====================================================================================================================================

TL,DR Not sure if the OP has the same issue I had, but I had black screen when playing games or just sometimes watching Youtube videos. The issue in my case was the HDMI cable.
The cable itself worked fine with other card but will sometimes cut-out when using Vega card. Swapping the cable to slightly shorter higher quality cable fixed my issue, ie no more intermittent black screen cut-out.

Not my issue but a good troubleshoot for anyone who comes across this thread at a later date
Im using a display port cable I got with my DELL screen which hasn't caused me any issues

 

[KLiKzg]

Thank you so much. The BIOS update was a bit nerve racking but has seemed to have helped as well as turnign down the RAM speed. No overclocking for me, I'm onyl plannign to undervolt the GPU for better temps, less power draw. Does that count as an OC? I heard that undervolts couldn't damage the card as you'd get a blue screen of death if you powered it too low

They could not damage the card, but it you could not get back to default setting - your GPU might be not working (aka bricked).

To remedy that issue, you would have to know how to reflash the BIOS on GPU, which also can permanently brick your GPU.

Are you ready to that step?

 

[Jackrkl]

They could not damage the card, but it you could not get back to default setting - your GPU might be not working (aka bricked).

To remedy that issue, you would have to know how to reflash the BIOS on GPU, which also can permanently brick your GPU.

Are you ready to that step?

Thanks for letting me know this. People always talked about it as free FPS for less power as thermal throttling is reduced which sounded like a no brainer.
Will defo be hesitant about proceeding