• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

3700x temps

R3X

R3X

Soldato
Joined
9 Aug 2013
Posts
3,553
AMD themselves said a pea sized dot in the centre or two peas or the cross method.

I went for the cross method myself but I use to do it via the thin applying with credit card, goodness knows for sure if it makes sense.

AMD vid was here shows it at the end:
https://www.youtube.com/watch?v=IP7HsNWSY4o

The reducing volts seems to have given good results but I think a few guys said it does reduce the performance a little. These 570 mobos and Ryzen chips are still fresh, I think it may still take some more bioses to get the volts and mhz fixed fully. However try the reseating see if you have any joy.
 
Associate
Joined
20 May 2009
Posts
1,053
Location
Essex
AMD themselves said a pea sized dot in the centre or two peas or the cross method.

I went for the cross method myself but I use to do it via the thin applying with credit card, goodness knows for sure if it makes sense.

AMD vid was here shows it at the end:
https://www.youtube.com/watch?v=IP7HsNWSY4o

The reducing volts seems to have given good results but I think a few guys said it does reduce the performance a little. These 570 mobos and Ryzen chips are still fresh, I think it may still take some more bioses to get the volts and mhz fixed fully. However try the reseating see if you have any joy.

I am curious to see if re-applying the paste has any sort of difference - in my experience it never has equated to much, but i'm fresh out of ideas on how best to find the balance. Under volting def keeps things cooler, but sure knocks the performance too.
 
Associate
Joined
28 Nov 2011
Posts
1,661
I am curious to see if re-applying the paste has any sort of difference - in my experience it never has equated to much, but i'm fresh out of ideas on how best to find the balance. Under volting def keeps things cooler, but sure knocks the performance too.

Noctua paste I'm using says to do one dot in the middle and then one in each corner.

I did that, obviously it sort of spread itself once the heatsink sat on and I finally managed to clamp the thing down.

3600 here with idle around 32-42c.
 

R3X

R3X

Soldato
Joined
9 Aug 2013
Posts
3,553
Seems to be the going rate for 3600/3700x cpus really, I think many are thinking its a 65watt cpu so should run much more lower temp but its not the case since its packing many cores and power under the hood.

I noticed these AMD cpus are sensitive to idle also, you only need a virus scanner or anything like any app running in the background for it to not really hit true idle cpu temps either, eventually you just get back to using the PC and ignoring it all once your happy of course with temps.

A good video to give you a heart attack on 'too much thermal paste or too little'

https://www.youtube.com/watch?v=EUWVVTY63hc
 

R3X

R3X

Soldato
Joined
9 Aug 2013
Posts
3,553
Also just a random question but is there really meant to be a big difference between top end air coolers from Noctua and Bequiet the £75+ range and top end top end Water coolers ?

I always thought it was on average 5-10c better, no idea since never watercooled

I know some reviews show top end air coolers giving liquid coolers a good run.
 

TrM

TrM

Associate
Joined
3 Jul 2019
Posts
744
Also just a random question but is there really meant to be a big difference between top end air coolers from Noctua and Bequiet the £75+ range and top end top end Water coolers ?

I always thought it was on average 5-10c better, no idea since never watercooled

I know some reviews show top end air coolers giving liquid coolers a good run.

I’ve gone from a Corsair h115i platinum to a Noctua nd-h15 chromex edition the whole ltt limited edition pushed me to get it :)

Not exactly cheap but paid 89.99 for it from Noctua official outlet and got to say I’m really impressed.

my h115i on my3900x under extream benchmarks was around 80-85 degrees and under full fan load the fans was noticeably noisy would t say as bad as older auto but defo noiseier then I would like

my Noctua on testing was about the same temps and very similar all core boost speed also but a lot quieter then my h115i was you can bearly hear it at in my phanteks elvov x case.

I also really like the look of a massive heat sink and this all black Noctua looks amazing altho with both fans on it doesn’t play nicely with my Corsair vengeance pro rgb the ram is to tall. But ordered some new ram to solve that problem and sold my Corsair stuff on eBay.

but at the end of the day the really big air coolers like dark rock pro and nocutua d15 will match a lot of aio out there and beat so many of the cheaper aio
 
Soldato
OP
Joined
27 Mar 2013
Posts
9,141
I’ve gone from a Corsair h115i platinum to a Noctua nd-h15 chromex edition the whole ltt limited edition pushed me to get it :)

Not exactly cheap but paid 89.99 for it from Noctua official outlet and got to say I’m really impressed.

my h115i on my3900x under extream benchmarks was around 80-85 degrees and under full fan load the fans was noticeably noisy would t say as bad as older auto but defo noiseier then I would like

my Noctua on testing was about the same temps and very similar all core boost speed also but a lot quieter then my h115i was you can bearly hear it at in my phanteks elvov x case.

I also really like the look of a massive heat sink and this all black Noctua looks amazing altho with both fans on it doesn’t play nicely with my Corsair vengeance pro rgb the ram is to tall. But ordered some new ram to solve that problem and sold my Corsair stuff on eBay.

but at the end of the day the really big air coolers like dark rock pro and nocutua d15 will match a lot of aio out there and beat so many of the cheaper aio
From what i've seen, due to the heat being generated in such a small area (rather than the bulk of the chip), it doesn't seem to make much difference to temps what cooler you use. FWIW i'm using the standard prism cooler on a 3700X and it seems ok. At idle im in the low 30's and if i startup something like battlefront 2 i average about 70. Benchmarking seems to go to 75-80, however i've set a custom fan profile to make it quiet.
 
Associate
Joined
24 Feb 2010
Posts
213
I’ve gone from a Corsair h115i platinum to a Noctua nd-h15 chromex edition the whole ltt limited edition pushed me to get it :)

Not exactly cheap but paid 89.99 for it from Noctua official outlet and got to say I’m really impressed.

my h115i on my3900x under extream benchmarks was around 80-85 degrees and under full fan load the fans was noticeably noisy would t say as bad as older auto but defo noiseier then I would like

my Noctua on testing was about the same temps and very similar all core boost speed also but a lot quieter then my h115i was you can bearly hear it at in my phanteks elvov x case.

I also really like the look of a massive heat sink and this all black Noctua looks amazing altho with both fans on it doesn’t play nicely with my Corsair vengeance pro rgb the ram is to tall. But ordered some new ram to solve that problem and sold my Corsair stuff on eBay.

but at the end of the day the really big air coolers like dark rock pro and nocutua d15 will match a lot of aio out there and beat so many of the cheaper aio

I'm using the D15 as well but I swapped out the front 140mm for the new A12x25 120mm to give me RAM clearance, using low ish speed fan curve, approx 800rpm under load I'm getting 59° in R20 and around 65° in Aida64 stress test.

XUlLxfG.jpg

XUlLxfG
 

R3X

R3X

Soldato
Joined
9 Aug 2013
Posts
3,553
I’ve gone from a Corsair h115i platinum to a Noctua nd-h15 chromex edition the whole ltt limited edition pushed me to get it :)

Not exactly cheap but paid 89.99 for it from Noctua official outlet and got to say I’m really impressed.

my h115i on my3900x under extream benchmarks was around 80-85 degrees and under full fan load the fans was noticeably noisy would t say as bad as older auto but defo noiseier then I would like

my Noctua on testing was about the same temps and very similar all core boost speed also but a lot quieter then my h115i was you can bearly hear it at in my phanteks elvov x case.

I also really like the look of a massive heat sink and this all black Noctua looks amazing altho with both fans on it doesn’t play nicely with my Corsair vengeance pro rgb the ram is to tall. But ordered some new ram to solve that problem and sold my Corsair stuff on eBay.

but at the end of the day the really big air coolers like dark rock pro and nocutua d15 will match a lot of aio out there and beat so many of the cheaper aio


Cheers, yeah always felt temps were not that far off or similar at least between AIO and high end coolers.

I prefer the big oversized coolers since no pump noise, extra electricity and can use the most silent fans without much issue.
 
Soldato
Joined
23 Nov 2007
Posts
4,939
Location
Lancashire, UK
Just got a 3700x so thought I'd read up on this as I've been having fan fluctuations...

Using a Noctua D15 with both fans.

- Full Linpack loading doesn't exceed 72C over 50 iterations at standard, so I'm happy with that.
- In Windows I regularly get fan fluctuations with typical idle temps around 40C but often spiking to 55C.
- If I fire up a game then the fan is stable at a relatively low RPM.

I think I'm going to just set a fan profile in the BIOS which is constant until 70C and then ramps right up, seems that however the temperature is being measured is too sensitive to a spike rather than an overall heat soak. I can see why people are making comments about whether the IHS is to blame, definitely the oddest thermal performance I've seen (I'd question my cooler/paste, but 72C at full load gives me confidence it's installed well).
 

TrM

TrM

Associate
Joined
3 Jul 2019
Posts
744
Just got a 3700x so thought I'd read up on this as I've been having fan fluctuations...

Using a Noctua D15 with both fans.

- Full Linpack loading doesn't exceed 72C over 50 iterations at standard, so I'm happy with that.
- In Windows I regularly get fan fluctuations with typical idle temps around 40C but often spiking to 55C.
- If I fire up a game then the fan is stable at a relatively low RPM.

I think I'm going to just set a fan profile in the BIOS which is constant until 70C and then ramps right up, seems that however the temperature is being measured is too sensitive to a spike rather than an overall heat soak. I can see why people are making comments about whether the IHS is to blame, definitely the oddest thermal performance I've seen (I'd question my cooler/paste, but 72C at full load gives me confidence it's installed well).

7nm small die and being off Center are another 2 reasons why heat spikes happen but all they are is heat spikes. 72 degrees at at full load is about right but instead of doing a constant fan profile till 70 change the delay of ramping fan speed. On my 3900x I have the normal fan curve but changed the delay to a lot longer that way if you cpu does stay Higher then fan speed can go up a little to help.
 
Caporegime
Joined
12 Jul 2007
Posts
40,511
Location
United Kingdom
7nm small die and being off Center are another 2 reasons why heat spikes happen but all they are is heat spikes. 72 degrees at at full load is about right but instead of doing a constant fan profile till 70 change the delay of ramping fan speed. On my 3900x I have the normal fan curve but changed the delay to a lot longer that way if you cpu does stay Higher then fan speed can go up a little to help.
Good suggestion that.

What exact setting did you use?

I am currently using a fan controller on the Wriath Prism with my 3900X and have lowered the SOC voltage down to 1.025v which reduced temperatures a little bit.

I have the fan speed set to a constant speed so its audible, but only just and its a constant noise rather than a ramping noise which you talked about.
 

TrM

TrM

Associate
Joined
3 Jul 2019
Posts
744
Good suggestion that.

What exact setting did you use?

I am currently using a fan controller on the Wriath Prism with my 3900X and have lowered the SOC voltage down to 1.025v which reduced temperatures a little bit.

I have the fan speed set to a constant speed so its audible, but only just and its a constant noise rather than a ramping noise which you talked about.

I use a Noctua d15 on my 3900x currently but I set the delay to 3.9 seconds before my fan speed changes whilst my fan speed does change sometimes but my own testing with ryzen master and watching the behaviour over a few days my idle temp spikes don’t last much of 3 seconds really and having the delay slightly above that means a lot less fan speed changes.
 
Soldato
Joined
3 Jan 2006
Posts
24,953
Location
Chadderton, Oldham
I'm running a Evo 212 cooler, can get audible slightly when temps go up, for gaming I'm hitting around 80 max, anything really taxing like cinebench = around 88 and Prime 95 = pegged at 95 degrees.
 
Soldato
Joined
1 Nov 2011
Posts
2,523
Location
Portsmouth
Thread bump - pretty much answered my questions, but is it worth me swapping out my Kraken X62 ?
NZXT'S Cam software is absolutely useless and finding an alternative to manage fan/pump speed is proving fruitless.

CPU hit 93°C running Heaven benchmark last night, and idles around 45°C - not comfortable with that.
 
Back
Top Bottom