B450 Tomahawk shutting down during gameplay after updating BIOS and CPU

general77155402dc

New member
Joined
May 22, 2021
Messages
21
After installing the latest chipset drivers direct from AMD (2.13.27.501) I have not witnessed a crash in over 5 hours of play time. Previously I would have experienced a crash before now so I'm optimistic. It would have been nice for MSI Center to offer the latest chipset drivers instead of something from February (2.11.26.106) as these drivers are clearly bugged. I was also able to get stability with the bugged driver 2.11.26.106 by disabling Precision boost overdrive and core performance boost in the BIOS settings so it seems like the instability may be related to these features.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
A few days ago I purchased a new 5900x to replace my 2600. The first thing I did when I got home was flash the bios to the newest version available - 7C02v1H5(Beta version). After flashing the bios I took everything apart, installed the new CPU and cooler, then reset the CMOS using the jumper cap. Leaving the bios settings at default, I booted into windows and started using the new CPU. General use is completely fine, as are benchmarking software. Other professional level software works fine without issue as well. Once I started playing a few games (Mount & Blade Bannerlord, Stellaris) is when everything started to fall apart.

After some time in game - ranging from 30 seconds to 4 hours, it seems to vary, the computer will crash; no blue screen, just black. This is reproduceable 100% of the time, I only have to load the game and wait to recreate the issue. Looking at the motherboard after this crash shows the CPU diagnostic light active. In addition, the motherboard seems to no longer be responsive as the button on the computer case no longer functions and the only way to turn the computer off is via the toggle on the power supply. I've tried everything I can think of to fix this without success but my computer seems disabled for games at this point after the BIOS and CPU update.

I have tried the following with no success:

1. Reinstalled Windows and all drivers (including motherboard via MSI Center)
2. Tested with a single memory stick in slot DIMMA2 (Corsair cmk16gx4m2b3200c16). I ran Memtest on this ram for 2 hours and no errors were identified so it appears to be fine. In addition, I’ve tested with the following ram settings:
auto settings (which defaulted to 2133mhz)
built in 3200 profile
manual timings that were indicated on the ram itself at 3200
tested with relaxed memory timings on the single ram stick at both 2133 and 3200
3. Tested an older set of G.Skill 2400mhz ram
4. Purchased a new Power supply as this seemed like a power issue (RM650x), ensured the CPU power and ATX power were secured, and made sure to split the power across two cables for my RTX 2070. I've even measured system load at the outlet with a meter while stressing CPU and GPU to 100% which resulted in a draw of 552W
5. Tested with no other PCI express cards installed
6. Tested with precision boost overdrive off
7. Updated the AMD chipset drivers to 2.17.25.506

Regardless of what I seem to try the computer continues to shut off. It seems like either the newest bios is faulty or the CPU itself is faulty. Are there known issues with this BIOS? Can I downgrade to another version that still supports my 5900x to test? I’ve spent $1200 on this upgrade so far only to receive a computer that can’t play games.
 
Last edited:

ryegrass

Active member
Joined
Feb 22, 2021
Messages
210
Check for CPU, GPU, RAM, etc. for overheating with a program such as HWMonitor.
You can also check Windows Event Viewer for errors which may give an indication of what's causing the crashes.

Capture.jpg

Also the best information I could find states that AGESA BIOS numbered 1.0.8.0 is the minimum necessary to post a Ryzen 5900x, since the previous bios on your board is 1.0.0.6, it doesn't look like you can revert to an earlier version and still use your 5900x.

https://www.pcworld.com/article/358...-motherboard-is-ready-for-amd-ryzen-5000.html
 
Last edited:

general77155402dc

New member
Joined
May 22, 2021
Messages
21
Check for CPU, GPU, RAM, etc. overheating with a program such as HWMonitor.
You can also check Windows Event Viewer for errors which may give an indication of what's causing the crashes.

View attachment 148848
Also the best information I could find states that AGESA BIOS numbered 1.0.8.0 is the minimum necessary to post a Ryzen 5900x, since the previous bios on your board is 1.0.0.6, it doesn't look like you can revert to an earlier version and still use your 5900x.

https://www.pcworld.com/article/358...-motherboard-is-ready-for-amd-ryzen-5000.html
CPU temp several seconds before the last crash was 65C while GPU was 57C. I'm not sure what ram would be but considering It was only running 1.2v or so when tested at 2133 I can't imagine it was running that hot. HWMonitor shows 45C from the motherboard sensors (PCH and MOS). Stellaris isn't a very demanding game so the temps never really get that high.

As far as windows event viewer there are no errors or warnings leading up to the crashes and because there is no bluescreen, the motherboard simply errors out, there are no crash dumps to check.
 

Alan J T

Well-known member
Joined
Sep 11, 2020
Messages
3,708
CPU temp several seconds before the last crash was 65C while GPU was 57C. I'm not sure what ram would be but considering It was only running 1.2v or so when tested at 2133 I can't imagine it was running that hot. HWMonitor shows 45C from the motherboard sensors (PCH and MOS). Stellaris isn't a very demanding game so the temps never really get that high.

As far as windows event viewer there are no errors or warnings leading up to the crashes and because there is no bluescreen, the motherboard simply errors out, there are no crash dumps to check.
650 watt PSU you never going to get that system stable with that small a PSU you CPU uses up to 200 watts GPU uses over 250 Motherboard about another 100 watts

I would be replacing the PSU with at least a 750 watt 80+gold or better to handle the power spikes that GPU and CPU can generate.
EDIT

I forgot to mention that the power spikes mainly happen in games and most time going from menu to game play or right after a loading screen while in the game. It seems to be linked to the fact the while in the menu or a load screen, GPU and CPU usage is very low below 100 watts and then ramps up to full power once game play resumes this is when the spikes can happen pushing the PSU past its normal operating range for a few microseconds causing the PSU protection circuitry to activate.
 
Last edited:

general77155402dc

New member
Joined
May 22, 2021
Messages
21
650 watt PSU you never going to get that system stable with that small a PSU you CPU uses up to 200 watts GPU uses over 250 Motherboard about another 100 watts

I would be replacing the PSU with at least a 750 watt 80+gold or better to handle the power spikes that GPU and CPU can generate.
EDIT

I forgot to mention that the power spikes mainly happen in games and most time going from menu to game play or right after a loading screen while in the game. It seems to be linked to the fact the while in the menu or a load screen, GPU and CPU usage is very low below 100 watts and then ramps up to full power once game play resumes this is when the spikes can happen pushing the PSU past its normal operating range for a few microseconds causing the PSU protection circuitry to activate.
I don't believe it's an issue of insufficient power. The computer under stress testing (P95 or CPUz stress) is only using a maximum of 280W as measured at the outlet with a meter (2.34A @ 120V). I've even gone so far as to run both 3Dmark and the CPUz stress test (which seems to draw more load than P95) simultaneously and even with both CPU and GPU locked at 100% the maximum current draw was 4.6A, or 552W of power at the outlet. This is well below the 648W available on the 12V rail.

CPUz and 3Dmark small.jpg


It's also important to note that in game I'm only seeing a draw of around 260W at the outlet so I'm nowhere near the limit in the games that show the crashes.
 
Last edited:

general77155402dc

New member
Joined
May 22, 2021
Messages
21

Alan J T

Well-known member
Joined
Sep 11, 2020
Messages
3,708
I don't believe it's an issue of insufficient power. The computer under stress testing (P95 or CPUz stress) is only using a maximum of 280W as measured at the outlet with a meter (2.34A @ 120V). I've even gone so far as to run both 3Dmark and the CPUz stress test (which seems to draw more load than P95) simultaneously and even with both CPU and GPU locked at 100% the maximum current draw was 4.6A, or 552W of power at the outlet. This is well below the 648W available on the 12V rail.

View attachment 148876

It's also important to note that in game I'm only seeing a draw of around 260W at the outlet so I'm nowhere near the limit in the games that show the crashes.
You need specialised equipment to measure the microspikes I believe it is the microspikes that are exceeding you PSU peak power that are causing the problem, they last for less than 40 microseconds, on average a 600 watt PSU can handle about 100 watts of peak power over the rated Max draw but only for a fraction of a second, depending on the PSU safety features it may shut down or disconnect the power after 10 to 40 microseconds each brand is different in how long and how much power they will allow the PSU to draw over the rated amount.

I never claimed you PSU was not big enough but that it can not handle the Power spikes caused by you hard ware while running games.

The info is hard to find for most PSU models some do have it or have been tested in lab conditions.

The PSU i use is a 850 watt 80+ Gold unit that has a peak power draw of 1050 watts for max of 40 microseconds so as long as a spike last lees than that it won't trip the protection.
And yes it took quite a bit of research before I found that info.
 

general77155402dc

New member
Joined
May 22, 2021
Messages
21
You need specialised equipment to measure the microspikes I believe it is the microspikes that are exceeding you PSU peak power that are causing the problem, they last for less than 40 microseconds, on average a 600 watt PSU can handle about 100 watts of peak power over the rated Max draw but only for a fraction of a second, depending on the PSU safety features it may shut down or disconnect the power after 10 to 40 microseconds each brand is different in how long and how much power they will allow the PSU to draw over the rated amount.

I never claimed you PSU was not big enough but that it can not handle the Power spikes caused by you hard ware while running games.

The info is hard to find for most PSU models some do have it or have been tested in lab conditions.

The PSU i use is a 850 watt 80+ Gold unit that has a peak power draw of 1050 watts for max of 40 microseconds so as long as a spike last lees than that it won't trip the protection.
And yes it took quite a bit of research before I found that info.
I can appreciate the research that went into discovering this; I believe the relevant metric for this would be Over Power Protection (OPP) which would trigger the PSU to shut down when the draw exceeds PSU capability? From the Tom's Hardware review of this supply the OPP seems to kick in at 807W, which is far beyond the ~260W the game us utilizing. If it is a micro spike issue it would mean the spike would need to cause a sudden surge of more than 300% power. I'd also expect to see this occur during stress testing where the system is using ~550W of power where even a much smaller spike could push it over the limit (though this smaller spike would still need to be ~150% the current power draw to trigger), however the crash never occurs under high load with these benchmarks, only in games that are far less taxing.
 
Last edited:

general77155402dc

New member
Joined
May 22, 2021
Messages
21
This issue may be resolved.

After installing the latest chipset drivers direct from AMD (2.13.27.501) I have not witnessed a crash in over 5 hours of play time. Previously I would have experienced a crash before now so I'm optimistic. It would have been nice for MSI Center to offer the latest chipset drivers instead of something from February (2.11.26.106) as these drivers are clearly bugged. I was also able to get stability with the bugged driver 2.11.26.106 by disabling Precision boost overdrive and core performance boost in the BIOS settings so it seems like the instability may be related to these features.
 

Alan J T

Well-known member
Joined
Sep 11, 2020
Messages
3,708
This issue may be resolved.

After installing the latest chipset drivers direct from AMD (2.13.27.501) I have not witnessed a crash in over 5 hours of play time. Previously I would have experienced a crash before now so I'm optimistic. It would have been nice for MSI Center to offer the latest chipset drivers instead of something from February (2.11.26.106) as these drivers are clearly bugged. I was also able to get stability with the bugged driver 2.11.26.106 by disabling Precision boost overdrive and core performance boost in the BIOS settings so it seems like the instability may be related to these features.
Can you run a CPU-z Bench mark and them validate and post the link please as it look like you CPU is underperforming. It should smoke mine by a good bit in CPU-Z benchmark stock bios setting but it is not for some reason.
This is my score with no Overclocking bios set to Auto on all Ram set to A-XMP 1 You cpu should be somewhat higher at least 1,000 points higher on Multi and score over 600 points on single thread
 

general77155402dc

New member
Joined
May 22, 2021
Messages
21
Can you run a CPU-z Bench mark and them validate and post the link please as it look like you CPU is underperforming. It should smoke mine by a good bit in CPU-Z benchmark stock bios setting but it is not for some reason.
This is my score with no Overclocking bios set to Auto on all Ram set to A-XMP 1 You cpu should be somewhat higher at least 1,000 points higher on Multi and score over 600 points on single thread
That bench was running both the CPUz stress test and 3Dmark at the same time in order to stress the whole system which is why the score was lower.

 

Alan J T

Well-known member
Joined
Sep 11, 2020
Messages
3,708
Thanks that shed a bit of light I noticed you're using dual monitors with totally different specks in Refresh rate and resolution. This can sometimes cause the problem you have been having if they are not using the same refresh rate and resolution.

And is it just in Steam games your having the problem I have noticed that some claim in the steam forums that apparently having Beta enabled can cause some games to crash. And since I have turned it off a couple of the games I was having some problems with crashing to Desktop have stopped doing so.
Screenshot 2021-05-26 102627.jpg
 

general77155402dc

New member
Joined
May 22, 2021
Messages
21
Thanks that shed a bit of light I noticed you're using dual monitors with totally different specks in Refresh rate and resolution. This can sometimes cause the problem you have been having if they are not using the same refresh rate and resolution.

And is it just in Steam games your having the problem I have noticed that some claim in the steam forums that apparently having Beta enabled can cause some games to crash. And since I have turned it off a couple of the games I was having some problems with crashing to Desktop have stopped doing so.
View attachment 148938
I was not participating in the beta and I've had dual monitors for years without incident. It was only after updating the BIOS and CPU that the crashes started. The crashes seem to be resolved for now after updating the chipset drivers but I'll repost in this thread if they return. I appreciate everyone's help with this.
 

general77155402dc

New member
Joined
May 22, 2021
Messages
21
Well, things seemed stable for a while but the crashing is still present, just less often.

I've tried updating to the newer 2.17.25.506 chipset driver but that hasn't solved the issue. I have also disabled PBO in the bios as I saw more stability on the older drivers with this feature off but it crashed with it disabled as well.

I'm looking for new suggestions on what it might be. Any help is appreciated.
 

ryegrass

Active member
Joined
Feb 22, 2021
Messages
210
Well, things seemed stable for a while but the crashing is still present, just less often.

I've tried updating to the newer 2.17.25.506 chipset driver but that hasn't solved the issue. I have also disabled PBO in the bios as I saw more stability on the older drivers with this feature off but it crashed with it disabled as well.

I'm looking for new suggestions on what it might be. Any help is appreciated.
It's been a while since your original post, so if you could provide more details that would certainly help:
Is the "crashing" happening at high CPU usage, idle, or when gaming, etc.?
Does it blue screen, randomly go to black and reboot, or something else?
Are there any Windows log errors when this happens?
Thanks.
 

general77155402dc

New member
Joined
May 22, 2021
Messages
21
It's been a while since your original post, so if you could provide more details that would certainly help:
Is the "crashing" happening at high CPU usage, idle, or when gaming, etc.?
Does it blue screen, randomly go to black and reboot, or something else?
Are there any Windows log errors when this happens?
Thanks.
It's the same behavior as previously mentioned. The crashing usually happens while gaming and low CPU usage; I've not seen a crash in applications other than games. There is no blue screen - the screen will suddenly go back and checking the motherboard will show the CPU fault light active. The motherboard at this point is no longer responsive as the computer power switch no longer functions and I have to toggle the PSU switch to shut off the computer. Since there is no bluescreen and the motherboard just seems to die there are no error logs in Windows event viewer - the computer doesn't even know it's been shut down as it doesn't even popup complaining "the computer wasn't properly shut down".
 

ryegrass

Active member
Joined
Feb 22, 2021
Messages
210
I know you and Alan J T already had quite a discussion about the power supply and it's certainly a good brand, but I think it needs to be eliminated as a possible cause. Do you have another supply you could use just to make sure that it's not causing the crashes.
 

general77155402dc

New member
Joined
May 22, 2021
Messages
21
I know you and Alan J T already had quite a discussion about the power supply and it's certainly a good brand, but I think it needs to be eliminated as a possible cause. Do you have another supply you could use just to make sure that it's not causing the crashes.
This was a new PSU I bought to replace my old corsair that was ~7 years old so this is the new "test" PSU. I don't have any other PSU beyond the old one that was crashing and this new one that is also crashing.
 

ryegrass

Active member
Joined
Feb 22, 2021
Messages
210
This was a new PSU I bought to replace my old corsair that was ~7 years old so this is the new "test" PSU. I don't have any other PSU beyond the old one that was crashing and this new one that is also crashing.
OK, I understand. What concerns me is that the computer doesn't just reboot which is what I've normally seen with other types of errors (RAM speed, PCIE bus, etc.) but stays off till you perform a hard reset with the power switch.

EDIT: Since you can't change out the power supply, there are still a couple of things to try:
Set PCH Gen Switch from [Auto] to [Gen3].
Try the motherboard outside the case on a non-conductive surface such as a wood table.
Remove and reseat the CPU.
 
Last edited:
Top