#Cpu clocks drop to 546 mhz under all core 100% loads
1 messages · Page 1 of 1 (latest)
Someone please help with this, haven't been able to figure it out for months
and you've been using it like that for months?
btw does it do the same when not 100% used?
idk like 80%
bios is new
good
i can take a video of it
games are fine
bios settings are all stock, just did a fresh install of win 11, was windows 10 before, same issue
all drivers are up to date
so if it runs games perfectly under 100% load
no games ever hit 100% load
ok
yeah its running perfectly
wdym
yeah
how long do you think youll be?
like 20-25 minutes
ok
PPT?
also did it start happening randomly?
hmmmmm
PPT is 76 W
under 100% load?
yes
ok
with these clock issues
ryzen 9 5950x
its base clock is 3.6 not 3.4
does the motherboard really supplies 105 watts to the cpu?
105 is TDP
wdum
ah i was wrong
?
i thought it was 3.6
yea but not dropping them to .5 ghz lol
reasonable
If it’s boosting down to 500mhz there’s 2 reasons. 1 I can’t remember but it’s a settings issue, the other is severe thermal throttling
Remedy is you’re severely thermal throttled or probably a full windows reinstall iirc
its not thermal throttling and i just reinstalled windows
I’ve seen before switch cpu and their aio busts a couple of times this year
What’s your temps when it’s 500mhz?
44 C
Have you ever under volted either cpu
all bios settings are completely default
50 C
Performance power plan set to allow 100% all the time?
idk what that means
the pc was exactly the same before i switched cpus tho, ever since i switched from 3700x to this one ive had this issue
Well you need to set xmp/docp back on jn the bios
xmp is on
You also need to in hwinfo let us know what ram voltage is
So the soc
That can throttle
Soc Voltage ( SVi2 tfn)?
Threads I see suggest stock 1.25v set by mobo on AM4 throttles
Should be around 1.4 to 1.45
After that i am not sure where you can go if it’s around that. If not need to manually adjust in bios
And xmp must be on or it just won’t load your ram (likely)
Did you change cpu lately?..
Did he install windows?..after changing cpu?..
Using a usb stick?..
Wiped the drive?..
I am thinking...the previous settings for the old cpu..in the windows or in bios may have affected the new cpu...
he said bios is stock
so he probably did a usb install
also bios are up to date
Needs a confirmation from him...
a desktop cpu can only switch to low clocks if..the power plan is messed up or previous oc settings or undervolting settings affected the new cpu...
Other than thermal issue...
I wonder if he is using a debloated os...
Looks bugged out. Check bios output
As 1 is too low
Default should be around 1.25 and xmp 1.4 or 1.45
Let me guess, you have a water cooling AIO?
how could that be a problem
Classical problem with water cooler (primarily AIO), they die, CPU heats up, overheats and downclocks to 0.5 GHz not to cook itself. One of several reasons not to buy any sort of water cooling over good air cooling.
but he said temps are 44
temps are fine
That's the reason his temps are "low" but frequency stays low
Since it now runs at so low speed it doesn't heat up enough even if the aio is fucked and simply dissipates though the coolers plate thermal mass alone.
according to him it slows down when on 100%
not sure if he check temps when the cpu is running on its max
as much as it can
Yes, because it ramps up, overheats, downclocks, cools down and stays there
Remember kids don't buy aios if you don't want to have random problems like that. 😄
i think fans are pretty good bruh
not a big fan of water cooling
personally
!!
The only thing AIO is good for is looks, performance is same as good air cooling and unlike air, it can be maintenance heavy and definitely not problem free
agreed
plus it increases costs
buying best once's available so that you wont regret in the future
Even Noctua D15 is cheaper than many AIOs and will be problem and maintenance free essentially forever.
@mental briar If HWiNFO is running at all times, what is the max temp recorded for the CPU?
He had all cores boosted momentarily to 4.9ghz
Max cpu die temp I can get from the video is 75c as it’s grainy
But it doesn’t thermal throttle like a dead aio would. A dead aio would let your cpu boost to TJ max before throttling
I’ve seen this issue before and various things have solved it. It’s never straight forward
It’s like his cores are parking but he’s never had an x3d chip
Actually… OP did you update motherboard chipset drivers?
Fully wiped drive did a full clean install, didn’t touch power plan,
AIO is new, and the temps are fine
It’s not a thermal issue
Yes they’re newest version
Sorry guys I went to sleep
Because its new and temps are low doesn´t mean it isn´t DOA or that it isn´t a thermal problem
It’s fine while gaming
It’s not a cooling problem, It’s been intermittent at times and I’ve had runs of cinebench going at max load at 70 C and then it starts
Plus if you watch the grainy video you can see that the second I click start on R23 the clocks die, it doesn’t even start the render before they all drop. There’s no temp spike as there’s no time
Let me guess, if you run another bench right after the first one fails, it always "fails" and throttles immediately.
Yeah, it always does it
Do yourself a favour and save time, go and grab even some basic cooler like Deepcool AK400 and try again
or even a box AMD one
What do you consider "high gaming load"
Because playing COD, Fortnite or whatever AAA fps you are trying isn´t high load for the CPU
High enough to bring temps up
If my cooler wasn’t working, it doesn’t matter what kind of load it would be it would still thermal throttle
It’s a 360 mm radiator with six fans
That doesn´t foking matter, the plate on the CPU can transfer X amount of heat and the water can take up some amount as well before its too hot to transfer any more out. If you just load CPU lightly, it might not be enough to overwhelm the cooler on the CPU and water would take potentially hours before enough heat would transfer.
GO AND GRAB EVEN A BOX COOLER AND TRY IT FFS.
Its 4 screws
Instead you are gonna spend hours trying software and refusing to admit that AIO can be DOA
I know how a cpu cooler works, I’m an engineering student I take courses on that shit
Is grabbing a screwdriver and changing a cooler to test it too difficult then?
I don’t have a second cooler atm, and also the cooler works
I’ll show you
Cold startup
Multi core, temps are good
Second pass
Temp is now 33 C
They are good only because the CPU throttles all they way to 500mhz and generates barely any heat man
I’m doing single core rn at 5 ghz, 59 C
Actually here let’s do more
Anyways I gotta go to work , if all you’re going to do is try and tell me it’s the cooler when it’s clearly not a thermal problem please stop responding to this thread as it’s not helpful
If you refuse to try things out when everything points in one direction then there is no helping you.
Indulge me, how do you know that the AIO works.
Also that’s not how heat transfer works. With a dead pump the CPU with throttle very quickly because as the water around the CPU block heats up the energy transfer rate slows down drastically. That is why you need a pump to cycle cold coolant through so that heat actually transfers. The whole water cooling loop does not heat up before your cpu throttles
Heavy single core loads, gaming temps are fine, temps stay stable in a hot room, I can hear the pump running, and like i said before I’ve had full cb runs with max temp at 72 C
I also see the reported pump speed in bios
Also the thermal throttling indicator in hwinfo stays at no
And the inlet hose on the cpu block is cold, not warm, meaning new water is being cycled through
And another thing, if it were thermal throttling you would hear the fans going crazy trying to cool it down
Apparently bit locker enabled can also cause this
Like I said it’s not an easy one to solve. Many issues can do this
It could also just be the motherboard and the VRMs failing though. It’s where I am hedging my bets
Maybe even knocked the board and a sensor dead when reseating cpu/aio
I thought it was motherboard vrms but idk how to prove it
Not sure if I have that, if it’s an app then no but if it’s a windows thing maybe
I’m always extremely careful when doing stuff like that, obviously nobody’s perfect but I don’t think that’s what happened
Where can I check this in bios? I’ll check when I’m home from work
Throw your old cpu back in and see if it throttles
Is my best guess
I sold it unfortunately, but I think it didn’t draw enough power to throttle the vrms, and now that I have a more powerful cpu they can’t keep up
You could try limiting the tdp of the cpu, and slowly raise it, until it doesnt work anymore.
Just a thought tho
How do I do that?
Is it in the bios?
AMD Ryzen Master. Its an app you can use, officially from AMD
Ah yeah I used to have it, didn’t like it
I’ll try that later tonight
I’m at work for the next eight hours
Thanks
where is the setting to limit tdp?
You can watch a youtube vid for rhat. I dont remeber as I aint have a amd cpu anymoew
@mental briar are you stress stesing both cpu and gpu?
No, just cpu
Thanks for the help, I know this is a super weird issue
well guess what
i am also learning new stuff
its interesting
can you try changing cpu drivers?
idk like just dont use the current one
use 3rd party app to either update them or use some older one
You mean motherboard chipset
CPU has no drivers
yeah ywah
any except the current one
not sure if it will work or not but worth the try
Okay I’ll try it
thats what i said
There’s a lot of reasons this seems to occur. Trying everything, everyone suggests is a good idea.
Yeah
I can’t seem to find an older chipset bios to install
Is there an archive website
you mean chipset drivers?
see idk where to find them
someone said they use iobite to update their drivers, so maybe try that?
ask dethriguez if he knows any better
Ok
Will do, I shouldn’t ttry to downgrade my bios right
I already have the latest drivers, we think they’re bugged and want to try an old version