Question I don't understand this.

[Deleted member 2720853]

Another cooling related issue I think?

Recently re-built my Ryzen build because my Gigabyte board died on me. This time I have an ASUS ROG Strix B550-F and Ryzen 9 3900X, with a NH-D15 as cooler, using NT-H2 thermal paste.

The way I applied the paste is by putting a dot in the middle (a little less when I put the motherboard in my PC after testbenching -- I put too much on my testbench and it oozed out of the cpu as I mounted the cooler). However even with a single dot, the paste DID spread over the entire IHS perfectly when I took off the cooler.

I tightened the cooler alternating between screws until both screws stopped and I couldn't tighten anymore. On the other build, I didn't tighten the cooler completely and I had reached 90C in AIDA64 and sitting over 75C in Cinebench. However on this build, I reached a max of 81C in AIDA, alternating between 79-81. In Cinebench it's around 71C max, but mostly sitting at 70C.

In gaming, it's sitting between 50-55, in the more CPU heavy games going to 65C.

What I don't get is that my boost clocks are the same. 4.02 GHz on stock.

What am I doing wrong? Why is the CPU not boosting as it should? I see people reaching over 4.2 GHz with PBO on this chip and I'm sitting here reaching 4.02 GHz.

When I apply my manual OC of 4.35 GHz on one CCD and 4.2 GHz on the other CCD, at 1.25V, the temps are better and the scores are higher.

But wht I don't get is why the stock CPU settings don't boost as high.

The only change I made to the CPU cooler fans is setting them to go 1100 RPM max because of the temp spikes on Ryzen making fans ramp up.

Did I mess up the cooler installation or paste installation again? What is the best way to apply NT-H2 to this CPU? I don't get it.

 

[Deleted member 2720853]

I still have the hardware. I also have a SECOND CX 650 that was also bought brand new, tried that PSU, same issues. No issues on the Intel build. I don't know what is causing this anymore.

I bought a NVMe for the ROG Strix build so yes, completely clean install.

I generally avoid beta BIOSes but I might give it a try I guess. Doubt it would help though.

Yes I have the recent B550 driver, 2.10.13.408, released 19th of October this year.

Those two drivers are also installed.

Only issue in Device Manager is the driver for the USB-C port on my GPU which doesn't exist, lol.

 

[Darkbreeze]

Try bumping your DRAM voltage by .005 to .020v, whatever the least increment it will allow you to adjust by is, in the BIOS. Save settings, exit. Test. If it still does it, do it again and raise it by another increment until you've raised it by .020v total. Some boards, especially Gigabyte, only allow .020v increment adjustments but your ASUS should allow .005v increments. If by .020v there is no change, then you might want to try and fiddle with the SOC voltage a bit.

This kind of problem is usually a RAM issue, or a PSU issue. Other than that, I'm at a loss as to what it could be man.

 

[Deleted member 2720853]

If it were a RAM issue I'd assume it would happen during overnight stress tests yet it doesn't.

It's definitely PSU.

 

[Darkbreeze]

No, there are a number of memory problems that won't occur under stress, and will ONLY happen when at idle or when resuming from idle.

For example. In Memtest86, test 10 which is the "bit fade test" tests to see if any bits have "faded" while sleeping or at lower power states. While probably not relevant to simply being at idle, it is still a type of problem that occurs while NOT under a load or actively addressing. Also, if there is a program or environment running with a memory hole problem, you could run out of memory even while at idle. These are of course only a couple of possible things that could happen, and they are not very common, but neither I think is your problem. So if the PSU doesn't cure the issue which I agree it probably might, but am somewhat skeptical since more than one power supply has resulted in the problem not being resolved, then I'd maybe start looking at these less likely scenarios.

 

[Deleted member 2720853]

... Don't you think a new board and memory modules would've nullified the possibility of a RAM issue?

I really don't wanna mess with SoC voltages.

 

[Deleted member 2720853]

I've already ran Memtest86 suspecting it's the issue and no luck, overnight tests with several passes completed successfully. Both with XMP on and off.

I spotted my CPU boosting to 4.65 GHz, this is with C-states on. I think it's boosting as it should?

Also, here is a screenshot of the Cinebench R20 single-threaded test, and HWiNFO sensors window. You can see the boost behavior of my CPU. Is this correct?

 

[Deleted member 2720853]

I also forgot to mention.

One time when I was testing C-states, the PC hard shut down, not reboot. Then when I booted the PC, the CPU light was active and won't go past it. I fixed it by turning my PSU off and on. If this is valuable info to determine what's going on.

 

[Deleted member 2720853]

Okay I apologize for the incessant posting, but I have to update here.

I increased DRAM voltage to 1.37v. I turned C-states back on, and set Power Supply Idle Control to Low Current.

Killed all programs in windows, left the CPU to idle.

All of these, should've shut down my PC in about a minute.

Now, all my cores in Ryzen Master are sleeping. CPU draws about 0.7W max. 0.20V voltage on idle.

I waited for a few minutes and it's still fine. Was this the problem all along? I hope so. I don't want it to shut down now. Please pray for me lol.

 

[Darkbreeze]

Isn't that what I recommended you should do? Seems like it is.

Try bumping your DRAM voltage by .005 to .020v, whatever the least increment it will allow you to adjust by is, in the BIOS. Save settings, exit. Test. If it still does it, do it again and raise it by another increment until you've raised it by .020v total. Some boards, especially Gigabyte, only allow .020v increment adjustments but your ASUS should allow .005v increments. If by .020v there is no change, then you might want to try and fiddle with the SOC voltage a bit.

This kind of problem is usually a RAM issue, or a PSU issue. Other than that, I'm at a loss as to what it could be man.

Hopefully that was it. I'll pray. LOL.

 

[Deleted member 2720853]

Well this is an improvement! Higher boost clocks on more cores, even up to 4.6, during single core test, and score went from 516 to 528.

Time to leave this running overnight and see if it still happens.

 

[Darkbreeze]

👍

 

[Deleted member 2720853]

👍

You sir are a wizard.

I left the PC on overnight. 9 hours passed and it still hasn't shut down, with no programs running.

Finally.

 

[Darkbreeze]

Nice man. Very glad it finally got sorted out.

 

[Deleted member 2720853]

But this is weird, why did it also happen on 3200 CL16 Vengeance LPX sticks?

 

[Darkbreeze]

Possibly a weak IMC on that CPU? IDK really, but this happens. Some CPUs are better than others. Some motherboards are better than others. Some memory kits are better than others. Even when they are all exactly the same part based on model number. It is, literally, the silicon lottery and it applies equally to all CPUs, GPUs and memory products. I can't say WHY bumping up the DRAM voltage works on some systems when another identical system works fine without doing that, but I can say that it IS the case from time to time.

 

[Deleted member 2720853]

Possibly a weak IMC on that CPU? IDK really, but this happens. Some CPUs are better than others. Some motherboards are better than others. Some memory kits are better than others. Even when they are all exactly the same part based on model number. It is, literally, the silicon lottery and it applies equally to all CPUs, GPUs and memory products. I can't say WHY bumping up the DRAM voltage works on some systems when another identical system works fine without doing that, but I can say that it IS the case from time to time.

Yeah. My integrated memory controller being weak and needing more voltage to the DRAM makes much more sense than two different brand new boards having the same issue even when the memory modules are swapped. After all, the CPU is one of the things that wasn't changed between builds.

Thanks again for helping me figure this out, it's been driving me crazy. Still going strong with no shutdowns so far.

 

[Darkbreeze]

That's great news man. Hopefully that was the resolution that was needed, and I think it probably was.

 

[Deleted member 2720853]

Lord, I just saw it boost to 4.75 GHz.

 

[Darkbreeze]

See, you should send me a 12 pack now. Nah, I'm just kidding. I'm truly glad it's working better now.

 

[Deleted member 2720853]

I just might! That's how big of an issue this was.

 

[Phaaze88]

To both of you.

 

[Deleted member 2720853]

Sigh...

No reboots, nothing seems unstable, CPU works just fine. Except these errors now exist.

I'll wait for the new BIOS to get out of beta and hope it fixes this. I'm apparently not the only one with these issues.

 

[Deleted member 2720853]

Increased SoC voltage to 1.125v. HWiNFO reports 1.106-1.119v. On Auto, HWiNFO reported 1.081v which might have been the problem. Let's see if this still happens.

 

[Deleted member 2720853]

So after increasing SoC voltage it shut down after an hour on idle. Reverted SoC voltage and it seems fine now, DRAM voltage still on 1.37v.

The WHEA errors are apparently caused by AGESA 1.1.0.0 when the Infinity Fabric clock is higher than 1600, many people are reporting this issue.

Guess it's the waiting game now. I hope these errors don't do anything detrimental.

 

[Darkbreeze]

Do you have PBO enabled or no? I can't remember? Might not matter anyhow, but you might want to see about adding a tiny bit of voltage via offset to the vcore (CPU core voltage) or however (manual) you want to do it.

I'm beginning to think, regardless that it is not common, that you might actually have a faulty or weak CPU.