Thread 714297042

>playing 10 year old dead online shooter with turbo mode off
>running 52 celsius, but getting too bad framerate
>decide to turn turbo back on with power plan (via registry hack) , instantly get more framerate, but temperature rises to 73 (limited from bios, would otherwise climb to 80+), amd gpu driver crashes and all of monitors turn black, game almost crashes too
What's the point of this shit? Why did AMD have to make this AI overclocker that keeps running temps hot? I have fancontroller (from razer) to control case fan speeds from software, and nothing I do seem to be able to tame the temps, putting the case fans to 80% or 20% bears very little difference. The paste has been changed like 2 or 3 times already in past few months, so I can't imagine it's the paste either. I can't imagine what could be going wrong for this thing to run so hot. Disabling the turbo decreased the CPU temp instantly from 73 to 53. And yes I'm undervolting this with that PBO2 tuner or some software, but it's still not enough to help tame these temps.

>>714297042 (OP)
they run at the temperature they should. don't worry about your old preconceptions of what temps cpus should run at. these are made to run at these temps. Like you I was surprised when I switched to the 5800x but thats just how they are meant to work

what's your cpu cooler
what's your case and how many fans do you have
guarantee this is user error

>>714297042 (OP)
Modern computers are, like everything else, designed to operate in a way that increases how often you replace them.

>>714297042 (OP)
??????

>>714297042 (OP)
>decide to turn turbo back on with power plan (via registry hack) , instantly get more framerate, but temperature rises to 73 (limited from bios, would otherwise climb to 80+)
This isn't hot at all. Modern computer hardware only starts thermally throttling around 95 degrees celsius, and shuts down around 100-105 to avoid reaching the temperatures where it can take real damage.

CPU temperature being around the 70s and 80s is no cause for alarm, panicking about that is not unlike 18th century retards panicking about the idea of being a car that moves faster than 50 miles an hour.

Got you covered bro.
Or get a Peerless Assassin for half the price for relatively same performance

>>714297359
its not the cooler. hes probably used to a cpu without PBO, these cpus SHOOT UP in temp during workload. Its not a defect, this is how they were designed to work.

>>714297042 (OP)
70 degrees is nothing for a modern cpu.

>>714297042 (OP)
I have a 7800x3d and my temps don't exceed 70 under load. I do have a 5 intake and 4 exhaust fans including the 360 rad, though.

>>714297042 (OP)
if you think 73-85 is too hot you're fucking retarded. they're designed to operate at these temps and dont start throttling themselves until they hit the mid 90s. Theres nothing wrong with these temps and if anything it sounds like you're using a shitty cooling solution as well.

>>714297042 (OP)
https://www.ocbase.com/download
tell us which of the tests fail for you

>>714297359
It's either noctua's nh-d15 or 14, has 2 fanse + 2 heatsinks. I have 1 exhaust cooler behind case (noctua ppc 3000), 2 front coolers (ppc 2000, some noctua 1500) and one top cooler (noctua 1500), all can be controlled from software. Case is Define 6 or something (the regular, not XL)

>>714297042 (OP)
>>decide to turn turbo back on with power plan (via registry hack) , instantly get more framerate, but temperature rises to 73 (limited from bios, would otherwise climb to 80+)
>amd gpu driver crashes and all of monitors turn black, game almost crashes too
These two issues are unrelated. Your CPU thermally throttling (it's not throttling by the way, 73 celsius is literally nothing to worry about) cannot make your GPU malfunction.

>>714297042 (OP)
it's not a GloFo shitter, it's fine at high temps

>>714297042 (OP)
i'm going to ask a silly question, anon
did you peel off the plastic film on the cooler?

>>714297603
>and my temps don't exceed 70 under load
While sustaining 5ghz? Mine hovers around 75C on an airflow setup on full boost.

Very surprised there's not one drawfag who has posted Op masturbating to his "too fucking hot" cpu

>>714297042 (OP)
temperature isn't heat you stupid fuck

>he's not using a mesh case
what's your excuse

>>714297042 (OP)
if you are taking money to push someones agenda at leas ask for some basic on-the-job training because you know shit about how this shit works

>>714297042 (OP)
thats not normal

>>714297042 (OP)
there is unironically something wrong with your computer
t. u12a

so how to reduce temps in geenral?
underclocking?

>>714297389
CPU-Z "stress test" is a joke
at least post cinebench temps
hard mode: OCCT or prime95 for extreme CPU assrape

>>714299183
undervolt

>>714299343
how?

>>714298253
holy shit you're a retard

>>714299385
Either with AMD software Ryzen Master, or with picrel

>>714299183
underclocking, getting a good cooler and better fans

>>714299183
Better cooler (ideally liquid) and strong case fans. Undervolt or get one of those AMD X3D series processors that uses like no energy.

does that proc have a Tctl reading?

>>714299965
>get one of those AMD X3D series processors that uses like no energy
Which one is that?

If you saw how retarded big modern heat sinks were, you’d understand the problem using modern software with the old ones

>>714300229
???

>>714297468
WTF, why have the second fan off center like that and not mounted over the ram? Is it less efficient?

>>714298253
If this is bait, you deserve a (You). I really fucking hope this is bait.

>>714297468
I think that's the same CPU cooler I have. Mine looks exactly like that

>>714299003
i think he might be onto something

>>714297247
Planned obsolescence

>>714297042 (OP)
After looking at other anons tests. You should check to see if the cooler is properly mounted and the heatsink is not clogged with dust, after that check to see if the thermal paste is actually spread out properly.

>>714297042 (OP)
71C is cool as fuck retard, Ryzen CPUs are rated up to 95C

>>714300780
I think it's possible my GPU (asus 7900xtx) is too big for the "small" define 6 case, since it's regular model define 6 and not the XL model, and it's possible that this large GPU is taking up air volume from the case (since running case coolers at high RPM doesn't seem to do any difference)

>>714300562
I had that same fan and I had issue with it cause BSOD when the fan touched the ram. My case was too thin to move it away so I ended up putting a smaller one on.

>>714297042 (OP)
you're supposed to adjust some bios setting to make it cooler and more performant

this is all the information I can give you because I forgot everything about it but it worked for me.

>>714301238
Well test it out by removing the GPU and see how it runs.

>>714300750
look you stupid cunt
do you understand that a giant bonfire and striking match can be at the same temperature
do you understand what heat is now that I dropped an example inside of your empty fucking skull

>Macbook Air M4
>fanless
>doesn't even get warm emulating PS2 games for extended periods of time

>>714299183
BIOS > PBO > curve optimizer > set to negative 15. Just werks and, unless you got a lemon, performs better because Ryzens, like all modern CPUs, are grossly undertuned so you end up with more thermal headroom without losing stability.

>>714297042 (OP)
Most fan software these days target a pretty high running temperature under load (like 80 - 95) to minimise fan speed (thus noise). The chip manufacturers say that's basically fine.

You can ofc always tweak it yourself to get better temps. But most people tend to just leave it on default.

>>714299303
Same temp

>>714301953
thx for the curve im steal it

>>714301731
>doesn't even get warm emulating PS2 games for extended periods of time
Wow, that's almost as good as a phone.

>>714301953
post final reaults dude. something is fucked up if you actually only get 900 ish

>>714301953
i hope you're not running those fans at 100% at the peak of the curve
the last 20% make them loud as fuck and more likely to break

>>714301692
Listen. I am going to take the bait again just this one last time JUST incase you are actually this stupid and I wish to uplift my fellow posters.

temperature
noun
tem·per·a·ture ˈtem-pər-ˌchu̇r ˈtem-pə- -p(ə-)rə-, -chər, -ˌtyu̇r, -ˌtu̇r
1
a
: degree of hotness or coldness measured on a definite scale
b
: the degree of heat that is natural to the body of a living being
c
: abnormally high body heat
running a temperature

Temperature is a measurement of thermal energy in a given environment. Its all relative to what is considered normal as to whether something is 'hot' or 'cold'. Temperature includes heat. You wont be getting anymore (You)s from me, so its better to find greener pastures.

>cpus are soldered on 300C+
>nooooo it's reaching 70C it's literally killing the cpu this is planned obsolescenc!

>>714302905
yes

>>714302905
death by a thousand cuts anon

>>714297468
PA is pretty based

It's because of the much higher energy density of these modern chips. A Vermeer CCX is the size of a pinkie nail and receives ~100W of electrical energy which all gets converted to thermal energy. But TSMC and AMD say that it's totally fine running this silicon at 90C and you don't need to worry about it.
The last line limit for PBO2 is simply the temperature limit, if you don't want your chip reaching 90C then you need to set a lower limit. The algorithm will always try to quench out the tiniest bit of clock rates until it hits the thermal limit.
Personally, I set my 5800X to 100W/100A/75C with a CO of -15 and it werks just as well. The only time it hits these limits is when I compile shit and then it still runs at 4.6GHz.

>>714297042 (OP)
>>running 52 celsius, but getting too bad framerate
fucking brown ESL retard

>>714297042 (OP)
But why do you care? Modern chips are made to run 24/7 in hot server environments. They can live long years running at 90ºC nonstop. My dad has an ancient Intel laptop where the CPU constantly reaches the low 90s and yet it's still functioning correctly.

>>714299343
> -25
That can't be stable. I started mine at -17 and now I'm at -12 because I keep finding workloads where the CPU fails and resets the system. Been at -12 for months now so I guess I've finally hit the perfect value.

>>714297468
phantom spirit is an option too

>>714303710
Sounds like you lost the lottery, most people run -20 fine.

>>714303710
It's a silicon lottery with the core stuff. I'm on a 7950X3D and found -10 to be stable but then I was reading around and some people were having no issues with -25 or something.

>>714303710
Mine is stable at -30, I just set it and forget it
Never crashed in like 18 months, including hours of stress tests

>>714303538
>But TSMC and AMD say that it's totally fine running this silicon at 90C and you don't need to worry about it.
i'm never drinking that koolaid
the 9800x3d has 20 transistors that take up the space of 1 2500k transistor and you're telling me they're just as stable? that shit has to be fragile as fuck

>>714297042 (OP)
Higher temps reduce cooling effort since the heat dissipates faster. Planned obsolescence might be a factor too, though

>>714303710
It depends on the CPU model. My other cpu can't run even -5 (has more cores, non-x3d). I think both 5700x3d and 5800x3d can run these higher - numbers (more undervolt) pretty well

>>714303938
>Never crashed in like 18 months, including hours of stress tests
Stress testing isn't the problem. Going idle is what usually made mine crash. Like waking up from sleep.

you're within temps. i don't know what you're doing that you think you're cpu bound. is this crashing in a testing suite or something? why does your gpu crashing (which you probably fucking up overclocking as well) have to do with your cpu? i just don't get you homie. also if this is you:
>>714297635
then you it seems like you have more invested in heatsinks and coolers (for a 5700x3d, really lol) than in your cpu, which is kind of funny.

>>714297042 (OP)
anon just opening a browser window with my 9950x3d jumps to like 100W of power and increases 10-15 celsius for a second

>>714303710
X3D chips can go lower because they aren't as aggressive with max boost clocks

>>714303991
>i'm never drinking that koolaid
Have you seen any of those CPUs die because of those temperatures?

>>714297042 (OP)
>but temperature rises to 73 (limited from bios, would otherwise climb to 80+)

how would I set a temperature limit on my CPU to do this?

t. dumbass who gets stressed out whenever any temps go over 80c

>>714303710
i run -25 and a 101 bclk

Bro still living in 2003 thinking his CPU will die if it gets too hot.
These things are smarter than you; they know to throttle when they need to. If it doesn't throttle, temps are good.

>>714304256
>anon just opening a browser window with my 9950x3d jumps to like 100W of power and increases 10-15 celsius for a second
You don't need to go that far. Open Ryzen Master. Let it idle for a few seconds. Now move the mouse.

>>714304264
i have killed my 3700x when i banged on the side of the case with a hammer to get rid of the ssd bay to fit my 9070xt
thankfully i was replacing it with a 5700x3d

>i underclock my cpu because it'll melt if i dont
It doesn't work like that, throttling exists for a reason.

>>714304496
What god forsaken case do you have?

>>714297042 (OP)
what the fuck is "turbo mode"? you should easily be able to set your all core curve optimization to -30, enable precision boost overdrive, and be done

>>714300562
bearing in mind that the open side of the fan is where the air comes in, these fans are mounted entirely incorrectly. they are facing the wrong direction and mounted in the wrong place as well. there are even arrows on the noctua fans that tell you exactly which direction the airflow moves.

>>714304523
Throttling isn't always on and clear cut. I've had my CPU reach 101°c even though it normally throttles at 97°c

>>714304702
Turbo mode is laptop cpus is what allow them to go up to their their max clock. Disabling it will force the processing to the base clock of usually 2.7-2.9Ghz. It's insanely good for keeping the temps low in gaming but woth some performance loss depending on games.

>>714304614
a shitty old thermaltake from my first actual build when i was still stupid enough to think i'd need an optical drive and several ssd and hdd bays
i could just replace it but idk if i can be bothered, the temps are fine without the side panel choking everything

>>714304117
I feel like I'm not getting any of the benefits from those case coolers though, for some reason. I can notice increasing CPU cooler speed or GPU cooler speed that my temp will drop, but making changes to case cooler speeds make almost no difference at all. My define case originally came with 3x fractal 1000rpm fans (2 intake fans front, 1 exhaust fan behind), and adding a single noctua 1500 on top of the case seemed to have done more to my coolnig that changing front fans to 2000 + 1500 or changing exhaust from 1000 to 3000. I haven't had chance to test with 2 top intakes that are both adjustable, but I noticed if I had 1 top intake that was spinning fixed 1000 or 1500 (depending if I put some non-adjustable noctua 1500 or non-adjustable fractal taht came with case), the temps went "all off", meaning having a top rear fan (an intake fan very close to the CPU/rear exhaust fan) was not good, so that one was disabled/removed very quickly in my testings. But basically, only the top-front fan seemed to have made any difference for me, having a cool 2000rpm front top + 1500 front bot didn't seem to help, neither the 3000ppc exhaust fan.

why do all prebuilt PCs have watercoolers now? they really want you replacing their shit every few years now huh?

>>714304274
You can do it from bios

AMD tells you to use liquid cooling.

>>714305378
i have a 7800x3d which will last me for like 10 years and i air cool it with no problems.

>>714297042 (OP)
That's the price you get for going <10nm
Max temperature on CPUs was ~60°C when it was 32nm

>>714305691
> use liquid cooler
> cpu boosts as far as it can go, hitting 95ºC again
Can't beat it without messing with BIOS settings.

>>714305984
How are smaller components more tolerant to high temperatures?

>>714305984
everything about this is incorrect
>>714306153
They're not

>>714304964
This is what I have done with the priority of dust prevention and so the case is positively pressurized. Do you find my CPU fans orientation retarded? Every arrow represents a fan.

>>714301486
I think I would try that, but removing the GPU on this mobo is risky. The release pin is very bad spot, and also I think you had to remove the CPU cooler to be able to remove the GPU. Basically it's ton of hassle. I'm thinking of driving my PC to some store and telling them to switch it into different a bigger case with more volume for air (will have to look up which case to go for). I honestly can't imagine it's anything other than case issue (which is a big issue though lol)

>>714306260
nta but I saw some youtube video mentioning that top exhaust might not be good, but I didn't watch it

>>714306260
no, this looks fine. reverse the two arrows over the cpu and you will have what was shown in the picture i initially referenced for example.

>>714303710
i'm running -30 on 5800x3d

>>714306260
Don't do this. The upper exausts in front of the cooler are stealing cold air and starving it.

>>714306260
>>714306386
Yeah the top exhaust will mess up the horizontal airflow and gimp the cooling performance. I believe noctua have some guidelines in their website.

>>714306153
the answer is really complicated and related to architecture, materials, fabrication and all sorts of things that have advanced significantly.
>>714306249
not intrinsically but our dies today are smaller and hotter. the reason they're hotter isn't because they're smaller though it's because of much higher power consumption.

i think its time for me to re-paste but im too lazy to

>>714305378
They don't? I bought a legion with a 13700KF and 4080 from lenovo back in 2023 and it had air cooling.

>>714297042 (OP)
What's the best bang for buck CPU at the moment? I'm thinking of upgrading this year but I mostly play pixelshit so I don't need to go crazy, but I still have an i5 from 2015 and it has shown its age in a few games I've played recently.

>>714306621
>the answer is really complicated and related to architecture, materials, fabrication and all sorts of things that have advanced significantly.
So it's _not_ because they are smaller?

>>714306153
Node shrinkage has required us to move towards FinFET and GAAFET tech which are actually much better structurally at dissipating heat.

>>714306653
> poor
9600x
> want more cores
9700x
> the best for gaming
9800x3d

>>714306648
I was thinking about buying a build with a 9700 XT and all of the CPUs had liquid cooling.

>>714297042 (OP)
>have R7 5700x3d
>my temp is 35C right now
>use Thermalright Peerless Assassin 120 SE
Get better CPU cooler.

>>714297042 (OP)
Enjoy your zeitgeist company

>>714306249
>Cram more shit in the same space = more power usage.
You don't need to be a genius to figure that one out.

>>714303402
I kinda wish I went for PA over the NH-D15. I did the D15 for the memes, but jesus fuck, you have to take that whole cooler off if you need to do anything with the PC after the fact that isn't just changing the GPU/storage.

>>714303402
that clock is terrible

>>714297042 (OP)
>buys a CPU designed to heat up to 90+C
>intentionally cripples it
some people simply shouldn't be allowed to own a computer

>>714306810
In practice, a modern AIO is going to last you 5-6 years. You shouldn't be using a tower cooler much longer than that anyway as it has the same problems (leak of internal liquid), and a wraith-style cooler just isn't sufficient for anything that's above 65W.

But if you're deadset on a no AIO - https://www.bestbuy.com/site/thermaltake-lcgs-view-i1497xs-350-gaming-desktop-intel-core-i7-14700f-32gb-memory-amd-radeon-rx-9070-xt-1tb-ssdnvme-m-2-snow/6625310.p?skuId=6625310

Thermaltake makes one with a standard tower cooler.

>>714306153
Smaller components means more components can fit in a die of the same size which will draw more power and generate more heat in order to work. And heat dissipation tech is barely keeping up.

>>714306857
>idle
ignored

>>714307086
I think 4150 is just single core boost, this is average across cores or something it still only goes up to 4050

>>714307193
Fine its 50C to 60C when it plays games.

>>714307086
shut the fuck up uninformed retard
it's a 5700x3d with 4.1ghz boost clocks
3.9 in occt is absolutely fine

>>714306805
Thanks, not poor but I buy pinball machines which consumes my gaming budget. Haven't decided just how much I'm willing to spend on a new PC yet.

>>714297042 (OP)
Unless your shit is running over 80 C then dont fucking worry about it.

>>714306284
I think you can unplug the power cable (if youre GPU requires direct power from the PSU) and still boot. Dont quote me on this but I have done this without issue.

>>714306386
>>714306601
>>714306607
I use the top 3 fans strictly to remove the heat that the GPU discharges upwards, I am also aware that the CPU coolers will be taking in GPU discharge air so I expect hotter CPU temps. Its not a problem for me since my GPU is the bottleneck for my gaming. I have tried with less fans up top, but it would create a heat pocket in the top right of the case and so I wanted that to be removed. Am I incorrect in this regard if GPU temps take priority?

>>714307343
cope harder faggot
its clearly untuned
a simple undervolt will boost those clocks.

> CPU thread
The same shit happens with GPUs. Am I meant to run my 9070XT with an hotspot of 95º and the VRAM hitting over 90ºC? This is meant to last 5-8 years running like this?

>>714307348
if you want to be a god, you go for 9950x3D, because 16cores + 3D (especially the 3D implementation of 9000 series) is killer combo

I remember when I got a GPU made within the last decade and thinking it was DoA since the fans weren't spinning at all for the first 30min. Then I learned they don't even turn those on until around 60C.

>>714307580
>1,13v
it's already undervolted you dumb nigger, stock 5700x3d runs over 1,2v in occt

>>714307681
I think modern PC components (especially when air cooling) require bigger cases. Buy XL model of whatever mid tower case, or some of the bigger size category Full Tower case

>>714307460
Best is to measure measure gpu temps after removing the top fans. Hot pocket shouldn't be an issue if no critical components are impacted

>>714308168
>require bigger cases
Doesn't make sense. Bigger case = more distance between the components and the case's fan.

>>714308235
bigger case means you have more volume of air you can juggle between components and coolers

>>714297042 (OP)
Why do you care what temperature they run at, provided it's within the design specifications and it works properly?
as for why:
>sensors become more accurate and precise, with more relevant temperature data being reported
>transistors become smaller and smaller and cores become denser

what mobo do you all have ?

re the CO shit: if you get WHEA/MCE errors or even outright crashes with just -15 then set "typical idle current" somewhere in EFI.
To run -30 you'll have to disable deep idle (C6 and I think C5 too). The crashes with -30 happen when it goes from deep idle to something like C1 or C0 immediately and this causes a voltage droop.
>>714307681
I had to limit my 7600 to the base 2300MHz speed because it kept trying to boost to 2600MHz for no fucking benefit in Expedition 33. I get practically the same performance/frame drops but at like 15C lower hotspot temps due to 50-60W less power usage. VRAM stays below 80C too due to the heat plate being able to soak up a bit more.
These small chips simply don't have enough cores to bounce around work while boosting, there's no chance for hotspot temps to drop in time for when boosting could actually do anything meaningful.

>>714308282
The name of the game when it comes to cooling a sealed box down that generates heat internally is air changes a hour. It does not matter the size of the box, what matter is how fast can you replace the air inside with cooler air. However, if the fans run into too much resistance due to, too small of an opening then getting a bigger space would help.

>>714308604
let me see the ports on the back. i love ports. the more the merrier

>sold my 12th gen to get 14th gen sinze z690 supports both
>14th gen has rust
>my 14th gen still works fine so im sticking with it
works for me. i dont care

>>714307460
>>714308174
Sorry not removing but turning first top fan into intake.

Here is the video this is based on
https://youtu.be/kdFQL3t5rmQ

>>714308602
>re the CO shit: if you get WHEA/MCE errors or even outright crashes with just -15 then set "typical idle current" somewhere in EFI.
oh, so that's why i never had any idle crashes
yeah this is good advice

>>714308757
very nice. loveas many fast USBs as possible. it'd be nice to have another usb-0c especially if it were 20Gbps (I'm assuming the oone you have is 10 as is mine). Also you have a fan and a 2 digit led on there so i can tell this was a serious/expensive motherboard. good job anon

>>714308856
Interesting. I might try that.

>>714297042 (OP)
Debunked. My dedktop APU is pretty cool. My Series S is pretty cool. My low end phone is pretty cool. Stop buying high end garbage

>>714297042 (OP)
lol, lmao

>>714297468
last 2 builds i just used aio, but if i ever go back to air coolers i would probably go with that monster

>>714308980
I think it was about £174 at the time, can't remember exactly.

>>714308646
oh so having bigger case wouldn't help me? My problem seems to be that despite having high rpm intake or exhaust case fans, they don't seem to be doing much, and I don't know why.

ngl i was a little scared it wouldnt pass in this heat. iv never seen my ram be above 50C before

>>714309523
sounds like it was a good deal.

>>714303710
I run -20 on CCD1 (non VCache) and -25 on CCD0 (VCache) of my 9950X3D.

>>714303858
I ran/run -30 all core with my 5800X3D.

>>714309759
Technically a bigger case would help regardless since there is more air volume in said space to suck up the heat. But it would be a minor impact with how much heat GPU dump out underload. A cramped case however might not allow the air to travel where it needs to go without creating a bunch of eddies which reduce performance and higher static pressure, meaning more resistance against the fans trying to push air in. Really depends on having good airflow channels, so if your set up does not really even have enough room to get your fingers to the motherboard. It could very well be a case issue. Make sure your fans are configured correctly and are facing the intended directions.

>>714308856
I got the idea for adding top front fan (intake) from watching some gamersnexus video where they tested some weird case, and they added a top front intake for some good results. seeing as I went from 1000rpm to 3000rpm with my exhaust fan, and there's almost no difference in temps whatever speed the exhaust is, having more exhaust doesn't seem to help me (I actually don't know what would at this point)

Alright lads
what app would you recommend to use for a full pc stress test? OCCT?
I got a shiny new 9060XT and i want to be 200% sure this thing is working as intended while i have my 14 day refund/replacement
It already driver crashed me while playing Wilds and bsod while idling but i'm blaming the fact that my windows install was so fucked that i had to DDU without safe mode
>>714308602
Is WHEA Cache Hierarchy Errors always related to cpu voltage? Because my brother pc with a ryzen 5 5600 is plagued by it, even replaced the cpu twice and the mobo but it still causes blackscreen reboots with no info about it other than the event log

>>714300562
I mounted mine a bit off center due to RAM. Works just fine.

>>714310772
>app

many modern cases are very cramped with these big air coolers, almost no room to do anything

>>714310772
I used several
>OCCT
>ycruncher
>prime95
>corecycler

>>714310152
coolercontrol looks cool
i wish there was something like this for windows, im so tired of bloat

>>714310772
>Is WHEA Cache Hierarchy Errors always related to cpu voltage?
In my case, it was always the negative curve. Updating the BIOS should be beinificial, too.

>>714311180
>tired of bloat
>asks for fan control that can be easily done through bios

Didn't peel of the plastic and is using a 65W cooler general

>>714297042 (OP)
Hope you finally killed yourself OP

>>714300562
>Is it less efficient?
Slightly, but these are already overkill unless you're overclocking.

>>714311180
fan control was the only thing that came close on windows side.
>>714311476
it cant. bios doesnt do hysteresis (well), or triggers, or controling arbitrary fans with arbitrary sensors.

I hope nvidia makes some cpu for gamers

>>714310954
anon, my brain is completely fried after trying to understand how my windows install was booting when it didn't have a bootmanager set up properly, it was so fucked up that in msconfig the boot tab was empty
>>714311137
I'll try these later, thanks
>>714311435
I'm pretty sure crashes started on stock setting and mostly during low usage/idle
i'll have to ask him again if it recently crashed since i remember manually setting some stock voltage/fabric memory settings a while ago

>>714297042 (OP)
It's only going to get worse. We've reached the cap on processor speeds with our current technology. We can only add more and more cores at this point to do more calculations in parallel giving the illusion of a speed increase, and that is exponentially increasing the amount of heat chips give off.

>>714313197
To add to this, the insane "stock" temperatures are because we're not getting faster instructions per cycle chips anymore, so to compensate these companies are basically overclocking their products from the factory. If you peel back the frequencies to say 99% max, it suddenly drops temperatures (and more importantly voltage) substantially. My 9950x3D can do 4.5Ghz all day with a peak temp of like 42C.

that sucks. I have 2 instances of Unreal Engine, an instance of Blender, and an instance of Photoshop open.... and I'm playing Elden Ring on max settings.

and my gpu is 42C and my cpu is 41C

sucks to be you, poor fag

>>714299183
modern mobos are cooking the cpus. default bios values used to be conservative, but today its more like full burn away. you can often undervolt by 30-40%, massively reducing tamps and without any performance loss
there was bit of affair when few literally blew up, but it didnt generate enough ruckus

>>714303538
>Personally, I set my 5800X to 100W/100A/75C with a CO of -15 and it werks just as well. The only time it hits these limits is when I compile shit and then it still runs at 4.6GHz.
I'm not familiar with touching this stuff. I've only set temp limit (from bios) and run the negative offset to apply downvolt value for PBO. Are there risks with reducing PPT, TDC or EDC ? I think I'm not shy about shedding some of the performance off (by manipulating the AMD algoritm so it's less aggressive), if it means I can run CPU cooler.

I have 5700x3d

How do I know if mine is getting too hot or not? What's considered too hot?

>>714309139
FX CPUs were large 32 nm chip and were actually easy to cool with good enough fan.

>>714316031
Anything above 60 is hot. I don't care for these defenders saying "hurr durr new electronics is made for 90c" bs. Nobody wants to be breathing those hot vapors. It's bad enough there's zoomers vaping nicotine, we don't need electronics vapor too.

>>714316031
60 or more is kinda warm for modern chips.

Is AMD really that much better than Intel?
t. 14700k enthusiast

literally download ryzen master and set curve to -20 (if you start seeing crashes just lower it slightly until stable)
9950x3d and running -20 gave me higher boosts for longer, and like 15c reduction in temps on a shitty peerless cooler

Post your motherboards

>>714316772
Hope you updated bios.

performance isn't free
you need aftermarket parts if you want to get more out of your stuff
or underclock it and get full life at the cost of performance

>>714316328
you're retarded arguing against physics because muh feefees

>>714317058
Are you AMD user? What do I do here >>714314331 ? Are there risks if I just set lower values to watt and amps ?

>>714316772
I am moving to AMD next CPU upgrade because fuck those E-Cores.

My 9800x3d seems to just boost itself until it hits 95 degrees when it's working near 100%. I was worried at first, but apparently that's just how they're made.

>>714305331
You installed at least one of the fans backward :)

>>714317327
Nope, they're rotating right direction

>>714316959
i did
>>714317250
explain? im on Windows 10 LTSC and everything runs smooth. i edit with premiere and everyon online says that intel quick sync still has the edge over AMD so i jeep buying incel. should i try AMD next time ?

>>714317186
Nope, I'm just stating the facts and not spewing this cope that amd/nvidia do. Efficiency used to be something to sought after, AMD used to get flak when their hardware ran poorer efficiency than Nvidia. We can't pretend it's not something worth seeking just because we've now hit a wall and keep pour more and more wattage to try raise benchmark numbers. Don't let corpos change your thinking like that, efficiency is everything.

I'm a retard, how am I supposed to undervolt my AMD CPU and GPU?

>>714317536
My desktop focus is gaming. I have an i7-12700k with 12 core (4 E-Cores) where the E-Cores are suppose to be used for windows background processes. So I really only have 8 cores. Developers fuck up and the E-cores get used for some games and causes stuttering issues. I would rather just have all of my cores be on the same level.

>>714318058
>Developers fuck up and the E-cores get used for some games and causes stuttering issues.
does windows 11's scheduler not fix that?

>>714318309
Yes, but I am on windows 10.

>>714318381
If you run power plan of 100% minimum and 100% maximum (basically max performance power plan), I think it gets rid of stuttering issues

>>714318381
shit me too.

>>714297042 (OP)
They also idle way too high.

>>714318620
>>714318626
Its not really an issue for me 95% of the time for most games. Its quite noticeable when I play some VR games like eaWRC 2024. Only other title I have had stuttering issues with not in VR is the recent oblivion remaster. Essentially any game with shitty CPU optimization. Luckily most games today are very GPU unoptimized.

>>714311137
for me its Space Engineers. its found instability in a cpu, my ram, and a gpu (memory specifically) on seperate occasions in different computers.

>>714318910
It's the memory controller and the NSA chip taking up the bulk of power when idling. The actual CCX idles a Watt or three when just displaying shit in a browser, which is actually pretty great.