Hot Temps for 5800X3D

[Mussels]

this thread is making me love my 5600x3d even more. don't even break 55 celsius in god of war 1440p ultra with fc140 cooler, and -25 all core.

it runs so cold!!!!!

Collecting all these tuning settings per series and collecting 'em all like pokemon is a current goal of mine.

Even basic mistakes people make like assuming BIOS settings for PBO are always active, when they can be disabled when PBO is or how PBO is totally ignored if static voltages are set trips up a lot of beginners.
(And how static voltages limit boost clocks, too)

 

[Space Lynx]

Collecting all these tuning settings per series and collecting 'em all like pokemon is a current goal of mine.

Even basic mistakes people make like assuming BIOS settings for PBO are always active, when they can be disabled when PBO is or how PBO is totally ignored if static voltages are set trips up a lot of beginners.
(And how static voltages limit boost clocks, too)

all I know is that my scores beat a 5600x in r23 and i dont get past 72 celsius or so, i do use a stronger fan curve than most people though, but meh, i love this chip. i also think i got a really good seal with my mx-6 and fx 140. more core to core variation on temps never exceeds 2 celsius from each other, which is mighty impressive based on my experience with Intel laptops.

 

[freeagent]

this thread is making me love my 5600x3d even more. don't even break 55 celsius in god of war 1440p ultra with fc140 cooler, and -25 all core.

it runs so cold!!!!!

I hear that man, in R23 I barely see 65c.

Running FC140 with a single fan.

 

[Space Lynx]

I hear that man, in R23 I barely see 65c.

Running FC140 with a single fan.

hmm i wonder how you get colder than me and have 2 more cores. i am running r23 again now and only seeing 61 celsius 2 minutes in. so yeah prob 65 before its all said and done.

not sure why i thought 72, might have been remembering wrong, i was just guessing as r23 i have not run in awhile. most benchmarking games last few weeks.

 

[freeagent]

hmm i wonder how you get colder than me and have 2 more cores. i am running r23 again now and only seeing 61 celsius 2 minutes in. so yeah prob 65 before its all said and done.

not sure why i thought 72, might have been remembering wrong, i was just guessing as r23 i have not run in awhile. most benchmarking games last few weeks.

Oh you let it run for 10 minutes?

Crazy. I could do that too I suppose.

 

[Mussels]

all I know is that my scores beat a 5600x in r23 and i dont get past 72 celsius or so, i do use a stronger fan curve than most people though, but meh, i love this chip. i also think i got a really good seal with my mx-6 and fx 140. more core to core variation on temps never exceeds 2 celsius from each other, which is mighty impressive based on my experience with Intel laptops.

Which is what i mean, sounds like you've dialed in some good settings. Go post em in the PBO settings thread here, so i can slap em in the OP

Smaller variations will happen just because we're all using different coolers, different thermal paste, different ambients... long term (30 minute) R23 testing is a good way to make sure it's comparable since that eventually evens out most smaller differences

 

[freeagent]

all I know is that my scores beat a 5600x in r23

What do you get in R23?

Edit:

freeagent`s Cinebench - R23 Multi Core with BenchMate score: 12635 cb with a Ryzen 5 5600X (hwbot.org)

freeagent`s Cinebench - R23 Single Core with BenchMate score: 1625 cb with a Ryzen 5 5600X (hwbot.org)

 

[AusWolf]

Old thread, I'm not sure if the original issue has been resolved or not, but in my experience, X3D doesn't mix well with water. A low-profile be quiet Shadow Rock LP achieved much, much better results on my 7800X3D than an otherwise fine 280 mm AIO.

 

[Mussels]

Old thread, I'm not sure if the original issue has been resolved or not, but in my experience, X3D doesn't mix well with water. A low-profile be quiet Shadow Rock LP achieved much, much better results on my 7800X3D than an otherwise fine 280 mm AIO.

I've seen similar with my dark rock slim.
Something about the baseplates they use is fantastic for AM4, and I guess AM5.

My DR slim outperformed all my 240mm AIOs, and temps were basically equal on custom water (worse, with GPU heat added in)

That said, I've tested the same 5800x on about 6 boards now and none of them had equal defaults so I'll always focus on PBO first - the second you pass the IHS/solder limits temps just skyrocket regardless of cooler.
The ITX system it's currently in runs at 70c with PBO enabled, and 90c with it disabled because Aorus don't understand what disabled means.

 

[Space Lynx]

What do you get in R23?

Edit:

freeagent`s Cinebench - R23 Multi Core with BenchMate score: 12635 cb with a Ryzen 5 5600X (hwbot.org)

freeagent`s Cinebench - R23 Single Core with BenchMate score: 1625 cb with a Ryzen 5 5600X (hwbot.org)

oh i guess i dont beat a 5600x. i got 11,419 multi 10 minutes.

Which is what i mean, sounds like you've dialed in some good settings. Go post em in the PBO settings thread here, so i can slap em in the OP

Smaller variations will happen just because we're all using different coolers, different thermal paste, different ambients... long term (30 minute) R23 testing is a good way to make sure it's comparable since that eventually evens out most smaller differences

thats whats strange, i dont do any settings changes. stock bios + xmp on ram + -25 all core + reboot

i keep life simple these days and it seems to work well enough, well with this rig it did. no coil whine and everything is running cold, im a happy camper

 

[freeagent]

That said,

I think you should try a nice Phantom Spirit 120SE on that 5800X, or X3D.. I think you would be mighty impressed..

As long as you dont have psychotic ambient temps lol

 

[tabascosauz]

I hear that man, in R23 I barely see 65c.

Running FC140 with a single fan.

stop giving me a heart attack man, I run 75C in R23 and was ready to rage until i saw 5600X

 

[Veseleil]

A low-profile be quiet Shadow Rock LP achieved much, much better results on my 7800X3D than an otherwise fine 280 mm AIO.

I've seen similar with my dark rock slim.
Something about the baseplates they use is fantastic for AM4, and I guess AM5.

Can confirm this, my BeQuiet performs unexpectedly great as well.

 

[Space Lynx]

Deepcool AK620 is on sale local store only for 37 bucks new. I was thinking of getting it to compare temps vs my fc140, gamersnexus rated it pretty good. I don't think i will do it, i still have a basically new in box 360mm aio i have no use for. Impusle buys... I'm getting better than i used to be. But still fail sometimes. So yea going to pass on it.

 

[tabascosauz]

Deepcool AK620 is on sale local store only for 37 bucks new. I was thinking of getting it to compare temps vs my fc140, gamersnexus rated it pretty good. I don't think i will do it, i still have a basically new in box 360mm aio i have no use for. Impusle buys... I'm getting better than i used to be. But still fail sometimes. So yea going to pass on it.

Isn't that kind of a steal? Usually a lot more expensive that's why usually PA120 is a no-brainer.

 

[AusWolf]

Can confirm this, my BeQuiet performs unexpectedly great as well.

Ironically, my 280 mm AIO is Be Quiet as well. It worked flawlessly with my previous CPUs, including an i7 11700 with unlocked power limits, but it craps itself on the X3D.

Now, I'm using a Dark Rock 4 that lets me run maximum boost all day long.

 

[Vya Domus]

Old thread, I'm not sure if the original issue has been resolved or not, but in my experience, X3D doesn't mix well with water. A low-profile be quiet Shadow Rock LP achieved much, much better results on my 7800X3D than an otherwise fine 280 mm AIO.

I don't think it has anything to do with being on water, it must be related to the cold plate design.

 

[AusWolf]

I don't think it has anything to do with being on water, it must be related to the cold plate design.

Probably. But that seems to be a bigger hit-and-miss on AIOs than on air coolers, which X3D chips are very sensitive for for some reason.

 

[Ka3el]

-30 simply may not be stable on your CPU
You haven't posted your system specs but RAM issues can cause that too - post a zentimings screenshot, at the very least.

With PBO disabled, the undervolt is disabled too. Sounds like you may not realize that some settings are disabling others

Hallo, so at least it seems you believe these "blackouts" can be caused by CPU unstable, thats good news
I use this RAM Crucial 16GB KIT DDR4 3200MHz CL16 Ballistix Sport LT Grey, I will download and postzentimings when I get home
The true is I didnt do any stability test, only gaming so -30 can be issue, but weird is this is really only happening when I touch power limits
PBO disabled? I dont know where other settings are, I didnt touch anything else than curve optimizer and power limits under PBO and I believe its active as before CO-30 BF2042 temps 90, after CO-30 temps 80

 

[Mussels]

oh i guess i dont beat a 5600x. i got 11,419 multi 10 minutes.



thats whats strange, i dont do any settings changes. stock bios + xmp on ram + -25 all core + reboot

i keep life simple these days and it seems to work well enough, well with this rig it did. no coil whine and everything is running cold, im a happy camper

You realise -25 is still a pretty serious undervolt? My 5800x at stock won't even do -5

Hallo, so at least it seems you believe these "blackouts" can be caused by CPU unstable, thats good news
I use this RAM Crucial 16GB KIT DDR4 3200MHz CL16 Ballistix Sport LT Grey, I will download and postzentimings when I get home
The true is I didnt do any stability test, only gaming so -30 can be issue, but weird is this is really only happening when I touch power limits
PBO disabled? I dont know where other settings are, I didnt touch anything else than curve optimizer and power limits under PBO and I believe its active as before CO-30 BF2042 temps 90, after CO-30 temps 80

As i said to Space Lynx, -30 is an extreme undervolt and load testing will not expose those issues - the errors happen at low loads.

You may not even have it active without touching those power limits, because you're not listing every setting you change i can only guess.
Enabled/auto/advanced on PBO are entirely different and not all will keep the undervolt active, if you touch the +MHz offset in any way, you're overclocking and undervolting and it's definitely not going to be stable.

 

[Space Lynx]

You realise -25 is still a pretty serious undervolt? My 5800x at stock won't even do -5


As i said to Space Lynx, -30 is an extreme undervolt and load testing will not expose those issues - the errors happen at low loads.

You may not even have it active without touching those power limits, because you're not listing every setting you change i can only guess.
Enabled/auto/advanced on PBO are entirely different and not all will keep the undervolt active, if you touch the +MHz offset in any way, you're overclocking and undervolting and it's definitely not going to be stable.

i guess it never clicked for me, but yeah i guess that is where the performance is coming from.

now I am cooking Greek food today, keep your stinky shellfish away for right now!

 

[freeagent]

Soft succulent lamb omg

 

[Mussels]

i guess it never clicked for me, but yeah i guess that is where the performance is coming from.

now I am cooking Greek food today, keep your stinky shellfish away for right now!

The 65W chips have their hardcoded limits lower than the higher end ones, so an undervolt will have a greater impact on them.

You could manage to stay under the stock PBO limits and be limited only by the maximum all-core and boost clocks, meaning it'll always stay at them - and that's going to make it perform a fair bit better than a stock system that might be hitting an EDC or PPT limit

 

[sfjuocekr]

The 5800X3D is simply put a hot chip, there is a waterblock sitting on my CPU that is cold to the touch and ambient is about 16 degrees. Under full load it will touch 80 degrees, but generally stays at 77-78.

Still touching the waterblock... when I stop the stress tests, the CPU dives back to 30-35 degrees immediately.

You should try to run a GPU and CPU stress test at the same time, this will increase the temperature of the CPU quite a bit

 

[Mussels]

The 5800X3D is simply put a hot chip

Hard to cool, not hot. Big difference.

Its the heat density, not the heat output - bigger coolers wont do a thing, but better thermal paste or better contact will change temperatures massively.