Pc not stable with xmp on. Help!

SpyderTracks

We love you Ukraine
Ohh i see what you mean. I thought we talking about the actual power option. No, hibernate is default. But i assume that the pc off and psu off that should not make a difference. I can turn it off and test. At this point im willing to try anything
If it’s hibernating and you’re powering it off at the PSU it’s never cleanly shutting down and the OS will gradually get more and more damaged.

Turning off the PSU isn’t really such a good idea on a modern machine, they’re not really designed for that kind of constant use, it’s only there to isolate the power but keep grounding when the system is being worked on.
 

Qwerty2236

Bronze Level Poster
Just to avoid any confusion. The only power options i keep are shut down and restart. When done, i shut down via start, once the system is off i turn the psu off also as i keep it covered when not in use and dont want that extra dust to settle on it. Also for someone to not accidentally turn it on when it is covered and. Hope it makes sense. In thia case i assume this IS a clean shutdown.
 

SpyderTracks

We love you Ukraine
ust to avoid any confusion. The only power options i keep are shut down and restart.
Yes, that's what I'm saying, that will hibernate rather than shutdown unless you've disabled hibernation in powershell. Shutdown in Windows = Hibernation.

So if you haven't disabled hibernation, your system will never have clean shut down, and the OS will be quite damaged as a result.

If you can run the command I asked, then we can see what's set, but it sounds like you still have hibernation active.
 

ubuysa

The BSOD Doctor
To keep thing simple; open an elevated command prompt and enter the following command...
Code:
powercfg -h off
That will definitely turn hibernation off, and with it Fast Startup.

I agree with @SpyderTracks 100% that turning off the power switch at the back is necessary - it's very hard to run a PC on accidentally. I'm also concerned that you cover it after shutdown. What that's doing is keeping the residual heat inside instead of allowing the components to cool down naturally. If you have so much dust around that it needs to be covered when not in use then it's in a way too dusty environment!

Please stop putting a cover on it at least, let the thing cool naturally. And if you're a little too hasty with the power switch at the back you can do some real damage to the OS.

BTW. I'm still concerned that you seem to still be trying to run that RAM at 6000MHz, despite PCS warning you that there will be known instability above 5600MHz.
 

Qwerty2236

Bronze Level Poster
Hopefully today i can try the hibernate off command and see. No pc is not in a dusty environment its just that i dislike dust. The psu off i had it since last time i had a desktop the psu sparked out of nowhere while the pc was off but psu was on. Plus the 4090 fiasco im a bit paranoid now. I know that pcs said the 5600 speed is what the system cpu can do. But seeing everyone around with xmp on why wouldn’t i try. If things just dont work then ill give up and manually dial down to a speed that it works.

At the moment it does work apart from the mentioned small inconvenience. If hibernate off does fix that then its all good.

And i dont cover the pc immediately after shutting down. Its around 10 mins after i quit the games it just idles before shutting down
 

SpyderTracks

We love you Ukraine
At the moment it does work apart from the mentioned small inconvenience. If hibernate off does fix that then its all good.
I'm not suggesting Hybernate off will fix it, I'm simply saying at the moment it's not completing a shut down, so the OS will be compromised over time.

Until we know the OS and software is healthy, we can't do any troubleshooting.
 

Qwerty2236

Bronze Level Poster
I'm not suggesting Hybernate off will fix it, I'm simply saying at the moment it's not completing a shut down, so the OS will be compromised over time.

Until we know the OS and software is healthy, we can't do any troubleshooting.
Here is the command
IMG_9631.jpeg
 

Qwerty2236

Bronze Level Poster
I have turned off hibernation and now it boots fine. It is all working perfectly… i hope i dont speak too soon but so far os great. Really appreciate ya’ll help!! Ill give it until next week to game this weekend and see how it is
 

SpyderTracks

We love you Ukraine
The psu off i had it since last time i had a desktop the psu sparked out of nowhere while the pc was off but psu was on. Plus the 4090 fiasco im a bit paranoid now.
The moral of this story is your doing out of spec things based on the above.

There is more to both these points:

PSU sparking out of nowhere. If the PC is actually turned off as in shut down, not in sleep, this isn’t possible on a properly rated PSU that’s on a surge protector. Would lead to ask which make and model PSU it was? But this is down to some kind of issue with the PSU being outside of safety regulations, or a surge on the line without a surge protector in place. Or if it was in sleep rather than being turned off, either due to a faulty PSU or a surge not protected by anti surge. Either way though, turning the PSU off at the PSU isn’t a way to protect against this. If you're simply saying you had a decent PSU fail in normal operation ie not while it was off, this is just how electronics are, and why warranties are so important. It's impossible to know each and every one out of the factory is fully 100% perfect, failures always happen as a percentage of components sold, again, nothing you can do to prevent failure other than being sure you've protected the system properly with an anti surge and proper cooling and airflow through the case.

The 4090 fiasco, I think you’ve entirely missed how that occurs. It can’t happen with the PC off as it’s receiving no voltage. It occurs due to the 12VHPWR connector not being fully inserted into the slot. Again, constantly turning the PSU off at the switch isn’t going to make any difference here.
 
Last edited:

SpyderTracks

We love you Ukraine
And also, just because it's working ok doesn't mean it's healthy.

Given that it's not been shutting down cleanly for an extended period of time, I would run the following commands in an elevated cmd prompt:

sfc /scannow
Dism /Online /Cleanup-Image /RestoreHealth

Be aware the DISM cmd could take quite a long time to complete depending on the OS degradation.
 

Qwerty2236

Bronze Level Poster
Well it has done it again. Just turning it on. But here is what i noticed:
IMG_9639.jpeg
IMG_9638.jpeg

Hwinfo says correct slot but quad channel and not dual channel. How is that possible?

They are in a2 and b2 a seen in the pic…
IMG_9640.jpeg
 
Last edited:

SpyderTracks

We love you Ukraine
I'll bow out, just not getting any response to suggestion posted, so it's impossible to know what's being done.

All the best with it.
 

Qwerty2236

Bronze Level Poster
I appreciate the help so far. Ill keep it short.

Everything on the mobo is default apart from the bios update that pcs sent to me.

I have the ram that’s specifically says 6000mhz. Runs fine at stock 4800mhz.

Just turning xmp on it makes the weird on off on again quickly but works fine with no crashes.

I reversed everything to stock enabled xmp and it beeps 5 times with tbe vha light on on the mobo. If i restart it boots fine no crashes no blue screens.

Again, nothing else has been done. Ram test is fine. Tried the single slot suggestion same thing happens with xmp on.

Its ok to not have any other suggestions ill close the thread then
 

Scott

Behold The Ford Mondeo
Moderator
In your shoes I would try playing with the timing. Well, actually I would have just sent the RAM back as it's not really stable/compatible. To get it working I would play with the timing.

30-36-36-76 is the current timing.

Try 32-38-38-96 to see if that's stable. If it is, job done. If it's not. Try down-clocking the RAM frequency to 5600 and see if that works. If it does, try the stock timings again to get a little more performance. If all OK still I would live with it, otherwise I would definitely send back.
 

Qwerty2236

Bronze Level Poster
Update:

I did
sfc /scannow
Dism /Online /Cleanup-Image /RestoreHealth

I tried the timings and failed. Pretty sure my fault

I reseated the ram, reset bios to default then only enabled xmp.

Worked. However i started to have random crashes 0xc0000005 (0).

Google points to some cpu errors for 13-14th gen cpu.

Did a mem test which is ongoing and this is what it shows. Mind you no errors with xmp off at 4800mhz.
IMG_9645.jpeg

Not sure if now its the ram or the cpu…
 

SpyderTracks

We love you Ukraine
I’m afraid the cooler is nowhere near suitable for the processor, neither is the case or PSU with the gpu. I’m not at all surprised it’s unstable.
I refer you back to what was said at the very start. Your configuration isn’t suitable for the processor. You don’t have enough headroom to overclock the CPU memory fabric which is what applying XMP does.

The processor is likely damaged due to overheating as a result.
 

Qwerty2236

Bronze Level Poster
Ok let’s say thats the case. Fine i can buy another cpu cooler. But even the gaming test im not going above 80c. Average temp say 50-65 c. If i had any temp issues sure hwinfo will tell me
 

SpyderTracks

We love you Ukraine
Ok let’s say thats the case. Fine i can buy another cpu cooler. But even the gaming test im not going above 80c. Average temp say 50-65 c. If i had any temp issues sure hwinfo will tell me
I keep repeating the same things over and over and you reply you understand and then repeat the same mistaken stuff, it's quite difficult.

You can't determine thermals with a game load, game loads go through spikes, it's not a stable load. But any game will result in 100% CPU spikes many times during gaming.

The only way to determine thermals is via a known load ie 100% with a stress test like Prim95 or Aida64
 

Qwerty2236

Bronze Level Poster
I keep repeating the same things over and over and you reply you understand and then repeat the same mistaken stuff, it's quite difficult.

You can't determine thermals with a game load, game loads go through spikes, it's not a stable load. But any game will result in 100% CPU spikes many times during gaming.

The only way to determine thermals is via a known load ie 100% with a stress test like Prim95 or Aida64
Ok will do it now
 

SpyderTracks

We love you Ukraine
Ok will do it now
You need to restart the pc, open only HWMonitor, then start the stress and leave it running for 10 minutes, while the test is still running take a snapshot of HWMonitor showing temps and frequencies across all cores on the cpu.
 
Top