[SOLVED] Ryzen 3600x on custom loop temp?

[Unick217]

Hello there!
Just a newbie question, im new in custom loop since ive been using cryorig r1 ultimate before, my temps in cryorig r1 ultimate seems similar to my new custom loop

Cryorig temps
Idle =45-50 c
Load = 75-78
Custom loop
Idle = 41-43c
Load = 75-78c

Note: this temps are from ryzen master, temps on aida64 that supports ryzen 3600x is a lot higher, like +5 to +8 c

Does my custom loop doesnt do any good than my cryorig?

Btw, im using bykski water block with 2 x 240 rads, and 2 res

Thanks all

Heres an image of my pc build

https://drive.google.com/file/d/1Sq4IFaBCpGyMGhThf4VoV7aDXNwncJt8/view?usp=drivesdk

 

[Flayed]

Hello there!
Just a newbie question, im new in custom loop since ive been using cryorig r1 ultimate before, my temps in cryorig r1 ultimate seems similar to my new custom loop

Cryorig temps
Idle =45-50 c
Load = 75-78
Custom loop
Idle = 41-43c
Load = 75-78c

Note: this temps are from ryzen master, temps on aida64 that supports ryzen 3600x is a lot higher, like +5 to +8 c

Does my custom loop doesnt do any good than my cryorig?

Btw, im using bykski water block with 2 x 240 rads, and 2 res

Thanks all

Heres an image of my pc build

https://drive.google.com/file/d/1Sq4IFaBCpGyMGhThf4VoV7aDXNwncJt8/view?usp=drivesdk

Very nice looking rig you have. Can't really comment on the temps, although they are alot lower than my 3600 with the stock cooler.

 

[Unick217]

Very nice looking rig you have. Can't really comment on the temps, although they are alot lower than my 3600 with the stock cooler.

Thanks man , im just worrying if somethings wrong with my water block mounting or something else, cause my custom loop doesnt make any difference from my past air cooler temp

 

[Phaaze88]

The temps are fine.

 

[Unick217]

The temps are fine.

Oh really?, it seems my air cooler then was on par with my custom loop now, btw im seeing other ryzen build having idle at 30c while mine is in 40c, anyway does it have to do with my environment? Im living in the Philippines

 

[Phaaze88]

There are various factors in place:
1)Liquid coolers have a slight edge over air at lower loads. Because liquid transfers heat more efficiently than air, the 'spikes' aren't as bad - you got better idle temps from it, but there's not much else to write home about.

2)Diminishing returns with overkill cooling.
The 3600X is very power efficient; it's not a hot and hungry chip compared to the likes of the 9700K or 9900K.
Neither the DRP4 nor custom loop you have would so much as break a sweat with this cpu.
A 50-60USD cooler would've produced similar results. Well, some models would've been a little louder while doing so.

3)Liquid cooling is ironically more dependent on chassis airflow than air coolers are.
The radiators themselves add another obstacle for air to pass through, or go around, thus reducing effective pressure. Case/rad fans need to be run higher to make up for that extra restriction.
I can't tell what the desk intakes look like - are they fairly open, or perhaps a small grille? It has to be taking in air from underneath where those radiators are.

4)I guess there's no vent beneath the power supply? Having it fan up was a little puzzling otherwise... but I guess it doesn't overheat.
I'm not too familiar with these desks.

5)I guess the gpu isn't running hot either? As vertical gpu in smaller chassis, it would be running rather warm.
I guess you also plan on adding the gpu to the loop later?

6)Room temps have a small effect on hardware temps. The chassis will have a larger effect.

All that aside, you put the cpu on a loop for aesthetics, right? And you'll do the same with the gpu?
If not, you wasted alot of money - well, you did get better idle temps...
Is it more quiet, or louder than the DRP4 is?

 

[Unick217]

There are various factors in place:
1)Liquid coolers have a slight edge over air at lower loads. Because liquid transfers heat more efficiently than air, the 'spikes' aren't as bad - you got better idle temps from it, but there's not much else to write home about.

2)Diminishing returns with overkill cooling.
The 3600X is very power efficient; it's not a hot and hungry chip compared to the likes of the 9700K or 9900K.
Neither the DRP4 nor custom loop you have would so much as break a sweat with this cpu.
A 50-60USD cooler would've produced similar results. Well, some models would've been a little louder while doing so.

3)Liquid cooling is ironically more dependent on chassis airflow than air coolers are.
The radiators themselves add another obstacle for air to pass through, or go around, thus reducing effective pressure. Case/rad fans need to be run higher to make up for that extra restriction.
I can't tell what the desk intakes look like - are they fairly open, or perhaps a small grille? It has to be taking in air from underneath where those radiators are.

4)I guess there's no vent beneath the power supply? Having it fan up was a little puzzling otherwise... but I guess it doesn't overheat.
I'm not too familiar with these desks.

5)I guess the gpu isn't running hot either? As vertical gpu in smaller chassis, it would be running rather warm.
I guess you also plan on adding the gpu to the loop later?

6)Room temps have a small effect on hardware temps. The chassis will have a larger effect.

All that aside, you put the cpu on a loop for aesthetics, right? And you'll do the same with the gpu?
If not, you wasted alot of money - well, you did get better idle temps...
Is it more quiet, or louder than the DRP4 is?

Yes, i have grilles underneath those rad, and all of exhaust are on the backside of the desk, the Power supply doesnt overheat as it would take some air from those fans on rad and vent on the back.

My gpu temps are fine nothings changed since my last pc desk build, it runs 60-75 c on games at ultra, btw im using these with 3 monitors surround, yes im planning to add loop on my gpu later and maybe another gpu?

I did use custom loop to lower my temps really, but i see it doesnt changed a lot, i thought it would be somewhere 30-40c at idle, it runs a little quiet than air cooling, im using 50% pump speed and auto fan curve in bios for my fan/rads. Using pump and fans at 100% doesnt do any justice either just a loud noise added, maybe ill try removing those grilles underneath my rads and see if i can lower my temps more, or maybe adding another rad?

 

[Phaaze88]

maybe adding another rad?

You're already past the point of diminishing returns. Adding another would make little difference.

maybe ill try removing those grilles underneath my rads and see if i can lower my temps more,

Might be worth a shot, but it would mean having to do maintenance more often.

Cooler capacity > pump speed > fan speed
A single 240mm rad is plenty for a 3600X. Past that and pump speed, it's fan speed in the pursuit of lower temps... at the cost of more noise.
Ehh, but they're fine as is anyway.

 

[Unick217]

You're already past the point of diminishing returns. Adding another would make little difference.


Might be worth a shot, but it would mean having to do maintenance more often.

Cooler capacity > pump speed > fan speed
A single 240mm rad is plenty for a 3600X. Past that and pump speed, it's fan speed in the pursuit of lower temps... at the cost of more noise.
Ehh, but they're fine as is anyway.

Oh, i see i think i get it now, maybe the temps difference between my cpu and other's loop was perhaps the IHS and the chip, i think i'll settle on this one, sure i did waste a lot of money but hey, my pc look awesome somehow lol, thanks man

 

[haseeb98ahm]

whats your clock speed? For Ryzen CPUs your temperature results are invalid if you test at stock speeds. Ryzen 3rd gen has Precision Boost 2 and XFR2. This means CPU keeps clocking higher across all cores as long as temps are below certain thrust hold.
With 3600x I saw no difference in temps on stock cooler and NH-D15. But when I manually set the clocks (4.35ghz @ 1.3375 volts) temps were about 10- 15 degrees cooler withNH-D15.

 

[Unick217]

whats your clock speed? For Ryzen CPUs your temperature results are invalid if you test at stock speeds. Ryzen 3rd gen has Precision Boost 2 and XFR2. This means CPU keeps clocking higher across all cores as long as temps are below certain thrust hold.
With 3600x I saw no difference in temps on stock cooler and NH-D15. But when I manually set the clocks (4.35ghz @ 1.3375 volts) temps were about 10- 15 degrees cooler withNH-D15.

Woah, do you mean my temps are a bit higher than it should be on custom loop?
Im running my cpu at stock speed, haven't OC'ed yet cause of that temp, should i manually clock my cpu too?

 

[haseeb98ahm]

I don't know if your temps are high or low for a custom loop, you can't check if your temps have improved unless you lock the multiplier. Rzyen 3rd gen CPUs boost using AMD Precision Boost 2 which is similar to how GPUs boost. In simplest terms, CPU auto overclocks as long as its below 75-80c. So, your temps will always be around 75-80c because CPU is boosting/overclocking until it hits 75-80c.

Note that I used the term overclock instead of boost because AMD's Precision Boost 2 is different compared to Intel's Turbo Boost 2.0. To keep it simple lets just say that Turbo Boost boosts some of its cores to Max boost speed and Precision Boost boosts all of CPU cores. And how close Precision Boost can get to max boost speed depends in the cooling used on the CPU.

 

[Karadjgne]

Temps, temps, temps. Bah. Custom loops aren't for trying to achieve the lowest possible temps, that's all on the efficiency of the cooler. Custom loops are for maintaining consistent temps across any workload on cpus that can't be cooled by regular means effectively.

Under thermal output values, liquid = air. At 65w output, a 140w CM hyper212 evo gets the exact same temps and performance as a Corsair H60. The only reason a Noctua NH-D15 gets better temps than many 240mm aios is the effectiveness and efficiency of the Noctua fans. Get to 250w and even the NH-D15 doesn't stand a beans chance vrs a 280mm aio.

Nothing magical about liquid cooling. A full custom loop has the ability to far exceed any aio cooling effectiveness and efficiency and thermal output, a decent 480mm can exceed 600w capacity. But if you don't have close to that output, it's cooling ability isn't any better than any other cooler.

Ryzen doesn't work temps the same as Intel. Intels lower core voltages and clocks over all cores, but workloads will use all the cores during background tasks, so loads and temps are shared. Ryzens shut down all the cores, using a single core to do all the background tasks, so has a significantly higher load on a single core. Idle temps upto the 50's are not uncommon, but that temp is on a single core, not all the cores, which would show low 30's on Intel.

Load temps are entirely different and are a representation of the ability, capacity and efficiency of the cooling vrs thermal output of the cpu, tempered by ambient temps.

 

[haseeb98ahm]

Temps, temps, temps. Bah. Custom loops aren't for trying to achieve the lowest possible temps, that's all on the efficiency of the cooler. Custom loops are for maintaining consistent temps across any workload on cpus that can't be cooled by regular means effectively.

Under thermal output values, liquid = air. At 65w output, a 140w CM hyper212 evo gets the exact same temps and performance as a Corsair H60. The only reason a Noctua NH-D15 gets better temps than many 240mm aios is the effectiveness and efficiency of the Noctua fans. Get to 250w and even the NH-D15 doesn't stand a beans chance vrs a 280mm aio.

Nothing magical about liquid cooling. A full custom loop has the ability to far exceed any aio cooling effectiveness and efficiency and thermal output, a decent 480mm can exceed 600w capacity. But if you don't have close to that output, it's cooling ability isn't any better than any other cooler.

Ryzen doesn't work temps the same as Intel. Intels lower core voltages and clocks over all cores, but workloads will use all the cores during background tasks, so loads and temps are shared. Ryzens shut down all the cores, using a single core to do all the background tasks, so has a significantly higher load on a single core. Idle temps upto the 50's are not uncommon, but that temp is on a single core, not all the cores, which would show low 30's on Intel.

Load temps are entirely different and are a representation of the ability, capacity and efficiency of the cooling vrs thermal output of the cpu, tempered by ambient temps.

This pretty much sums up the confusion with ryzen temps. We are so used to intel temps we assume AMD temps are slimier. AMD measure temps differently and it manages it cores and power differently.

Ryzen doesn't work temps the same as Intel. Intels lower core voltages and clocks over all cores, but workloads will use all the cores during background tasks, so loads and temps are shared. Ryzens shut down all the cores, using a single core to do all the background tasks, so has a significantly higher load on a single core. Idle temps upto the 50's are not uncommon, but that temp is on a single core, not all the cores, which would show low 30's on Intel.

^this is why AMD is so power efficient but idles at higher temps and AMD precision boost 2 is why load temp are high.

If you are gaming leave the CPU at stock for best single threaded performance.
If your workloads are mainly multi-threaded then overclock. you will sacrifice single threaded performance for much better multi-threaded performance.

 

[Phaaze88]

But when I manually set the clocks (4.35ghz @ 1.3375 volts)

You REALLY shouldn't leave it like that: View: https://www.reddit.com/r/Amd/comments/ejgc6p/1325v_is_not_safe_for_zen_2/?ref_source=embed&ref=share

Put the frequency to auto, or even PBO if you wish, just remove that OC, before you end up like some of the folks in that thread who hyper-degraded and destroyed their cpus...
No one actually knows what settings are safe to use, other than one's respective cpu.

The trick to maxing performance on Ryzen 3000 is:
-good cooling; the chips are temperature sensitive
-memory frequency up to 3733
-tightened memory timings(there are Ryzen memory guides out there)
The gains from all core OC are inferior to memory tweaking, and you don't have to sacrifice single core performance to do it.

 

[haseeb98ahm]

You REALLY shouldn't leave it like that: View: https://www.reddit.com/r/Amd/comments/ejgc6p/1325v_is_not_safe_for_zen_2/?ref_source=embed&ref=share

Put the frequency to auto, or even PBO if you wish, just remove that OC, before you end up like some of the folks in that thread who hyper-degraded and destroyed their cpus...
No one actually knows what settings are safe to use, other than one's respective cpu.

The trick to maxing performance on Ryzen 3000 is:
-good cooling; the chips are temperature sensitive
-memory frequency up to 3733
-tightened memory timings(there are Ryzen memory guides out there)
The gains from all core OC are inferior to memory tweaking, and you don't have to sacrifice single core performance to do it.

didn't know that, thanks for the info. I never ran OC long term cause of poor gaming performance.