Hot Temps for 5800X3D

[Mussels]

can assure you that this photo is just a false optical illusion.

It's straight from their website and no optical illusion. Anything with direct heatpipes with those gaps will perform worse on ryzen, vs even a weaker cooler with a solid baseplate.

These are low wattage CPU's when gaming, so they run really cool there - the only time they can get hot is something like all core AVX load in cinebench R23, and that depends on your PBO settings and what it chose to limit the watttage to.

 

[19Romulus84]

@Mussels Man, do you understand what I'm saying to you? This is just a photo that has nothing to do with reality. I have a similar cooler and have built several computers on this cooler and it has no gaps. Do you understand or do I have to repeat it to you 10 times? The foot of this cooler is perfectly smooth, like a baby's butt, and this is just a photo and a visual effect that has nothing to do with reality. You're stuck to it like velcro on a dog's tail and you're floundering. It's just a photo, badly done, some graphic effect, whatever. In fact, this foot is perfectly smooth. So stop writing this crap about some cracks because there aren't any cracks.

 

[freeagent]

You need to relax my friend.

Sometimes coolers go through multiple revisions throughout their cycle. You may have a smooth one, but the older ones might not be like that. I have seen it multiple times.

 

[19Romulus84]

@freeagent I doubt that's what's happening. Mine is over 4 years old and literally 3 days ago I was cleaning it because I was changing the processor and it was the same as in March 2019 when I bought it. This photo is a photo of a new cooler from the store and not tortured for 20 years. This photo just doesn't reflect reality, that's how weird this photo came out and that's it, it has nothing to do with reality.

 

[Mussels]

@freeagent I doubt that's what's happening. Mine is over 4 years old and literally 3 days ago I was cleaning it because I was changing the processor and it was the same as in March 2019 when I bought it. This photo is a photo of a new cooler from the store and not tortured for 20 years. This photo just doesn't reflect reality, that's how weird this photo came out and that's it, it has nothing to do with reality.

Like how people get confused with the cooler master CM212, it seems you don't realise how many revisions of those coolers exist.
If its got the plain old heatpipes like that, they don't perform well. It's a budget design.

Higher heat density means that smaller imperfections are a problem on these CPUs - the cooler can be big enough mass wise, have the right materials but perform like ass because you've got 3x to 10x the airgap above one of your cores.

 

[19Romulus84]

@Mussels All in all, if it were so, of course you're right, each such gap above the core caused bad heat dissipation, which would have a fatal effect on the temperature. I don't have any gaps, not even the smallest gaps. I have good temperatures in games 99% of the time 75-85 degrees. There are short, random moments once every 20-30 minutes of playing that literally for 10 seconds it will jump to 90% and then it will drop. This processor is like that, it works very well for me on a regular Fera 3 and I didn't do any undervolting or change anything. The processor itself is running by default and works very well.

 

[freeagent]

It looks like the cooler is having a hard time with it. 75-85 on the CPU just playing a game is pretty toasty.

 

[dgianstefani]

It looks like the cooler is having a hard time with it. 75-85 on the CPU just playing a game is pretty toasty.

Completely normal for 5800X3D, always runs hot due to the cache layer being another barrier for heat.

You won't fix this without delidding or a very high end custom water system.

 

[freeagent]

Completely normal for 5800X3D, always runs hot due to the cache layer being another barrier for heat.

You won't fix this without delidding or a very high end custom water system.

I only see those temps if I run WCG, with no fan on the cooler.

R23 only makes my CPU hit 65.. and I get over 15K.. 10c more if I run it with no fan..

 

[dgianstefani]

I only see those temps if I run WCG, with no fan on the cooler.

R23 only makes my CPU hit 65.. and I get over 15K.. 10c more if I run it with no fan..

Sure. Maybe you got a lucky sample or have a dodgy temperature sensor.

Most reviews, and user experience report 75 C+ under load, including TPU testing.

 

[freeagent]

Sure. Maybe you got a lucky sample or have a dodgy temperature sensor.

Most reviews, and user experience report 75 C+ under load, including TPU testing.

Maybe.

Clocks don't drop until temps get up there, who knows. Shouldn't need crazy water to cool the CPU though, it only does 143w PPT under specific conditions.

 

[dgianstefani]

Maybe.

Clocks don't drop until temps get up there, who knows. Shouldn't need crazy water to cool the CPU though, it only does 143w PPT under specific conditions.

It's not about the sustained load potential or higher wattage dissipation, but rather the better delta temperatures good watercooling can give, due to reservoir having a good heat buffer and water temp and thus coldplate temperature being very low compared to the baseplate of an air cooler or AIO. This helps mitigate the otherwise poor thermal conductivity of the setup.

It's not the wattage that causes the high temps with Zen X3D, it's the thick IHS and layers of cache etc. Why you see higher wattage Intel parts get lower temperatures on the same cooling setup.

These are my gaming temps, because I have a large reservoir, despite only having a 240/40mm radiator. Sometimes the GPU goes to low 50s.

 

[freeagent]

It's not about the sustained load potential or higher wattage dissipation, but rather the better delta temperatures good watercooling can give, due to reservoir having a good heat buffer and water temp and thus coldplate temperature being very low compared to the baseplate of an air cooler or AIO. This helps mitigate the otherwise poor thermal conductivity of the setup.

It's not the wattage that causes the high temps with Zen X3D, it's the thick IHS and layers of cache etc. Why you see higher wattage intel parts get lower temperatures on the same cooling setup.

I understand all of that, I am curious why mine is so easy to cool now.

I used to cool ~250-300 old school Intel watts with air

 

[dgianstefani]

I understand all of that, I am curious why mine is so easy to cool now.

I used to cool ~250-300 old school Intel watts with air

Maybe you just have a great VF curve capable chip. It happens. Some are harder to cool but can hit higher frequencies, some are much easier to cool but will never OC well. At least with 5800X3D you can't OC anyway so efficiency bins are more valuable.

I remember watching some professional overclocker, maybe der8auer, talk about how leakier voltage chips generally OC better.

 

[Sptz]

It's pretty hot here in the UK and my home office gets mega toasty. I can definitely see it in my build, 5800x3d -30 CO, thermalright PA120 with Arctic P12 Max fans on it, even with fans at 100% (N O I S Y) on cinebench it's reaching 80C.

I'm curious if overtime as well the cooler seating can start revealing hotpots? I can see some cores are quite hotter than others.

What is everyone's recommendation on PPT TDC and EDC? I've never touched those, should I also attempt an undervolt offset? No idea where to go here.

(I'll also reseat the cooler when I have the patience to clean up the thermal paste lol)

 

[dgianstefani]

It's pretty hot here in the UK and my home office gets mega toasty. I can definitely see it in my build, 5800x3d -30 CO, thermalright PA120 with Arctic P12 Max fans on it, even with fans at 100% (N O I S Y) on cinebench it's reaching 80C.

I'm curious if overtime as well the cooler seating can start revealing hotpots? I can see some cores are quite hotter than others.

What is everyone's recommendation on PPT TDC and EDC? I've never touched those, should I also attempt an undervolt offset? No idea where to go here.

(I'll also reseat the cooler when I have the patience to clean up the thermal paste lol)

You can use curve optimizer, but it's a bit of hassle.

The only real advantage of it is you may get slightly higher average clocks and slightly lower wattage. As for heat output it's always pretty low with averages around 50 W when gaming, peak around 140 W as with most single CCD Zen 3. It's always going to run hot unless you do extreme things like delid, but the performance advantage is minimal, maybe 25 MHz.

Cooler mounting etc isn't as critical as on an Intel system for example, as the issue is IHS/3DVCache related, the bottleneck is thermal transfer from the silicon to the cooler.

You could get slightly better temps with offset mounting, lapping, liquid metal etc.

https://www.overclock.net/threads/5800x3d-owners.1798046/ this can help if you want to tweak, but the X3D chips are as out of the box good as they come. BLCK OC, lapping etc, all come with risks. You won't even see the same RAM tuning performance bonuses that you do with normal Zen, since the 3DVCache already mostly fixes the latency issue, any 3600/16 or better will do fine.

With mine I just use Conductonaut Extreme liquid metal between the IHS and waterblock, that gives me ~5 C better temps than paste, but it's not really necessary.

A simpler solution would be to tune your fan curve setup using a combo algorithm, using Fan Control https://github.com/Rem0o/FanControl.Releases.

I do mine on a curve based off the average of CPU, GPU and coolant.

 

[Sptz]

You can use curve optimizer, but it's a bit of hassle.

The only real advantage of it is you may get slightly higher average clocks and slightly lower wattage. As for heat output it's always pretty low with averages around 50 W when gaming, peak around 140 W as with most single CCD Zen 3. It's always going to run hot unless you do extreme things like delid, but the performance advantage is minimal, maybe 25 MHz.

Cooler mounting etc isn't as critical as on an Intel system for example, as the issue is IHS/3DVCache related, the bottleneck is thermal transfer from the silicon to the cooler.

You could get slightly better temps with offset mounting, lapping, liquid metal etc.

https://www.overclock.net/threads/5800x3d-owners.1798046/ this can help if you want to tweak, but the X3D chips are as out of the box good as they come. BLCK OC, lapping etc, all come with risks. You won't even see the same RAM tuning performance bonuses that you do with normal Zen, since the 3DVCache already mostly fixes the latency issue, any 3600/16 or better will do fine.

With mine I just use Conductonaut Extreme liquid metal between the IHS and waterblock, that gives me ~5 C better temps than paste, but it's not really necessary.

A simpler solution would be to tune your fan curve setup using a combo algorithm, using Fan Control https://github.com/Rem0o/FanControl.Releases.

Thanks. Yeah, it used to never go above 70C (cinebench) just 2 months ago, now it's scorching. I'm guessing messing with power limits (PPT etc) isn't worth it then?

 

[dgianstefani]

Thanks. Yeah, it used to never go above 70C (cinebench) just 2 months ago, now it's scorching. I'm guessing messing with power limits (PPT etc) isn't worth it then?

Not really, it's the most efficient CPU for gaming besides Raptor Lake and Zen 4 X3D already. The issue is temperature delta between coolant medium (in your case air temperature) and the IHS. You can get slightly better temps by increasing the delta as much as possible, but it's not a wattage issue, it's a thermal bottleneck issue.

 

[Sptz]

Not really, it's the most efficient CPU for gaming besides Raptor Lake and Zen 4 X3D already. The issue is temperature delta between coolant medium (in your case air temperature) and the IHS. You can get slightly better temps by increasing the delta as much as possible, but it's not a wattage issue, it's a thermal bottleneck issue.

Thanks. The only problem I'm seeing is that it's not reaching 4450 all core on C23 now because of temps being at 79-80C but hey ho, such is life I guess!

I've also noticed that running a 30 min bench on C23 crashed at around 11mins, that crash that shows a popup dialog with a log.txt. Not actual hard lock or anything, no WHEAs. But this would usually be related to a too aggressive CO right? Though I've done countless corecycle stress tests, combined with P95 etc, and wasted hours to do a per core CO and couldn't get any core to crash at any CO level. But today it just decided to crash cinebench out of nowhere, I'm assuming this is heat related as it may be warming up VRMs etc?

 

[dgianstefani]

Unlikely to be VRM related as the board has reasonable VRM setup and it's by no means a high wattage CPU, plus high VRM temp won't crash the CPU, it will just degrade lifetime of the motherboard.

With Zen it's most likely memory/infinity fabric related, especially if your voltages are undervolted as you have.

There's no real reason to run -30 CO for the X3D, and it's unlikely that it's stable on all cores.

 

[Sptz]

Unlikely to be VRM related as the board has reasonable VRM setup and it's by no means a high wattage CPU, plus high VRM temp won't crash the CPU, it will just degrade lifetime of the motherboard.

With Zen it's most likely memory/infinity fabric related, especially if your voltages are undervolted as you have.

There's no real reason to run -30 CO for the X3D, and it's unlikely that it's stable on all cores.

I couldn't get it to boost to 4450 before running a negative CO. Maybe I can back it off to like -25 and see what happens. Could be my memory got it at 3600C14, tight timings but I've ran memtest 1usmus etc for over 24 hours to test stability

 

[Mussels]

Completely normal for 5800X3D, always runs hot due to the cache layer being another barrier for heat.

You won't fix this without delidding or a very high end custom water system.

not while gaming it isnt, thats R23 load temps
I see 45-60C gaming

W1zz's 5800x3D review only saw 77C on the CPU and 71c gaming in the newer 13900K review, at stock settings on air - because they're incredibly low wattage while gaming, since the datas reused from the cache

My CPU gets heated up by my GPU, and its cooler than that - heck my 5800x runs colder on air than it ever did in this loop for the same reason. They don't care about large coolers, only about the ambient temp (and a loop with water at 45C might as well be air in 45c ambients) and the contact quality.


Full size dual 4K doesnt uplolad to TPU, so cropped it is
I'll get a few screenshots, borderlands 3:
Yes this is undervolted and on custom water cooling, which is why i've been claiming higher temps are normal than this
(And the extra SoC voltage raises temps, not lowers)

and a realworld photo to show ambients arent exactly freezing cold, either - loops warm, since i've been playing borderlands1 all morning with @JC316

 

[dgianstefani]

not while gaming it isnt, thats R23 load temps
I see 45-60C gaming

W1zz's 5800x3D review only saw 77C on the CPU and 71c gaming in the newer 13900K review, at stock settings on air - because they're incredibly low wattage while gaming, since the datas reused from the cache

View attachment 300602View attachment 300605


My CPU gets heated up by my GPU, and its cooler than that - heck my 5800x runs colder on air than it ever did in this loop for the same reason. They don't care about large coolers, only about the ambient temp (and a loop with water at 45C might as well be air in 45c ambients) and the contact quality.


Full size dual 4K doesnt uplolad to TPU, so cropped it is
I'll get a few screenshots, borderlands 3:
Yes this is undervolted and on custom water cooling, which is why i've been claiming higher temps are normal than this
(And the extra SoC voltage raises temps, not lowers)
View attachment 300599
View attachment 300598

and a realworld photo to show ambients arent exactly freezing cold, either - loops warm, since i've been playing borderlands1 all morning with @JC316
View attachment 300601

Newer 5800X3Ds aren't as good as launch ones for temps. Regardless, 75 C instead of 70 C I don't see an issue.

This has been noted by YouTubers who've bought more than one, e.g. OptimumTech.

I assume this is because the best dies are going to server chips.

I also have a fairly early production 5800X3D, I bought mine two months after release. Recent production X3D chips are binned, the most efficient bins seem to be reserved for Epyc chips.

It's not something to worry about though ~ 60 C or 80 C is practically meaningless for a desktop CPU, especially one that can't be overclocked, you may use 1-2 W extra from voltage/leakage scaling due to the temperature, and maybe 25 MHz from boost scaling, but it's irrelevant, these aren't KS series chips you're trying to eke out 6.4 GHz instead of 6.3, where every °C matters.

Typical gaming load looks something like this, currently temperatures where I live are 20-25 C. The GPU is not bottlenecked, it's running a 236 FPS frame cap in NVControlpanel, and the CPU frequency is averaged, since gaming doesn't use all cores. My 240mm radiator manages to cool 400 W with no issues.

@Mussels if that RAM is B-Die you can get your tRFC down by quite a lot, besides absolute latency, tRFC latency is one of the biggest Zen performance variables for gaming, lower = hotter though, so ensure you have good airflow over the DIMMs. Mine have been repadded with 15 W/mK pads I had leftover from the GPU repad. You could probably get sub 300ns without even changing voltages. It's one of the easiest methods to tune for performance with RAM, because you can simply use this chart, and adjust your values by one step then test, until you become unstable.

For every 10trfc drop you get .1ns lower latency on AMD.

500trfc vs BDie 250trfc is 2.5ns lower latency just from tRFC value.

If you don't have BDie you'll struggle to get lower than 400 ns though.

I've got 190 ns which is decent, but there's guys out there running 150 ns lol.

I'm very curious to see how my temps change when I apply Conductonaut Extreme on both CPU/GPU, and upgrade GPU pads to 20 W/mK ones. Currently only have Conductonaut (original) on the CPU.

 

[Sptz]

Newer 5800X3Ds aren't as good as launch ones for temps. Regardless, 75 C instead of 70 C I don't see an issue.

This has been noted by YouTubers who've bought more than one, e.g. OptimumTech.

I assume this is because the best dies are going to server chips.

I also have a fairly early production 5800X3D, I bought mine two months after release. Recent production X3D chips are binned, the most efficient bins seem to be reserved for Epyc chips.

It's not something to worry about though ~ 60 C or 80 C is practically meaningless for a desktop CPU, especially one that can't be overclocked, you may use 1-2 W extra from voltage/leakage scaling due to the temperature, and maybe 25 MHz from boost scaling, but it's irrelevant, these aren't KS series chips you're trying to eke out 6.4 GHz instead of 6.3, where every °C matters.

Typical gaming load looks something like this, currently temperatures where I live are 20-25 C. The GPU is not bottlenecked, it's running a 236 FPS frame cap in NVControlpanel, and the CPU frequency is averaged, since gaming doesn't use all cores. My 240mm radiator manages to cool 400 W with no issues.

View attachment 300671

View attachment 300661View attachment 300662

@Mussels if that RAM is B-Die you can get your tRFC down by quite a lot, besides absolute latency, tRFC latency is one of the biggest Zen performance variables for gaming, lower = hotter though, so ensure you have good airflow over the DIMMs. Mine have been repadded with 15 W/mK pads I had leftover from the GPU repad. You could probably get sub 300ns without even changing voltages. It's one of the easiest methods to tune for performance with RAM, because you can simply use this chart, and adjust your values by one step then test, until you become unstable.
View attachment 300665

For every 10trfc drop you get .1ns lower latency on AMD.

500trfc vs BDie 250trfc is 2.5ns lower latency just from tRFC value.

If you don't have BDie you'll struggle to get lower than 400 ns though.

I've got 190 ns which is decent, but there's guys out there running 150 ns lol.

I'm very curious to see how my temps change when I apply Conductonaut Extreme on both CPU/GPU, and upgrade GPU pads to 20 W/mK ones. Currently only have Conductonaut (original) on the CPU.

What sticks are those? I've got these https://www.overclockers.co.uk/team...00c16-3600mhz-dual-channel-kit-my-002-8p.html

Absolute crap. Can't even get 3600C14. Only XMP actually works. 3600C14 worked for a few weeks, passed memtests for 2 days, corecyclers etc, then it started borking out... Had to run 1.47v (actual 1.5v on my Gigabyte b550i as it always overvolts) for this which is pretty bad, spent over a month trying to tune these only to end up at the beginning :l

 

[P4-630]

W1zz's 5800x3D review only saw 77C on the CPU and 71c gaming

Why don't I see the CPU temperatures when gaming on that page?

I only see the "Load" temperatures on that page...