water temps not very good

[Sayón]

You're wrong. Empirical tests I linked to show exactly what the difference is between thermal pastes when tested properly.

Again, it is MORE than obvious you had other variables going on to cause such a significant drop.

Please move on and stop spreading misinformation. THis is not helpful to the OP or anyone else reading it.

I'm only posting what I experienced with a non-standard thermal paste in combination with my hardware.
If there are other variables for this, (it can be) then it must be something from my mobo because all bios settings except ram timings, are all set to auto. (meg x399 creation)
If you are not open to others that's fine but don't argue something what someone else experienced and shares with others to (possibly) help them out.
End of discussion and peace to you to bro.

 

[EarthDog]

I'm only posting what I experienced with a non-standard thermal paste in combination with my hardware.
If there are other variables for this, (it can be) then it must be something from my mobo because all bios settings except ram timings, are all set to auto. (meg x399 creation)
If you are not open to others that's fine but don't argue something what someone else experienced and shares with others to (possibly) help them out.
End of discussion and peace to you to bro.

All I am trying to do is expose you to what actually caused such a dramatic difference. Is it me who isn't being open or you in this case?

I believe what you say happened did happen, however, the reasons behind the change have little to do with the thermal paste. I'm basing my posts on facts from empirical tests. If you were using mashed potatos as your thermal paste, perhaps I can see this...but otherwise, you need to understand that MX-4 is just middle of the pack and isn't the issue here... same with your application method. Your anecdote/help isn't helpful in this situation. It is misleading at best.

Let's be clear here for all reading this............ MX-4 is a mediocre paste...works just fine....but it is nowhere close to 28C drop fine, PERIOD.

....sorry peter...

 

[Elysium]

You're not really giving us enough data, we need graphs and charts, hard numbers giving us a look at what is happening when and where. If you don't have them already, get GPU/CPU-Z and HWiNFO and post a bunch of data shots. We need to see idle and load temps for your components as well as fan speeds. I kinda agree with Vya Domus regarding your original fan temps but, idk man, two rads well in excess of 300mm and two reservoirs?? I'd expect to see lower temps more in line with thesmokingman's rig tbh.

 

[Sayón]

All I am trying to do is expose you to what actually caused such a dramatic difference. Is it me who isn't being open or you in this case?

I believe what you say happened did happen, however, the reasons behind the change have little to do with the thermal paste. I'm basing my posts on facts from empirical tests. If you were using mashed potatos as your thermal paste, perhaps I can see this...but otherwise, you need to understand that MX-4 is just middle of the pack and isn't the issue here... same with your application method. Your anecdote/help isn't helpful in this situation. It is misleading at best.

Let's be clear here for all reading this............ MX-4 is a mediocre paste...works just fine....but it is nowhere close to 28C drop fine, PERIOD.

....sorry peter...

Ok, let me put it this way then..for me, that thermal paste did a wonderfull job. Call it luck or whatever it was, I'm more than happy with it so over and out. Anyways thanks for your response

 

[EarthDog]

Ok, let me put it this way then..for me, that thermal paste did a wonderfull job. Call it luck or whatever it was, I'm more than happy with it so over and out. Anyways thanks for your response

You can put it any way you want... but that doesn't mean the thermal paste was in any way a large part of that temperature drop or is it 'amazing' (like setting your ram manually, lol). Move on, Sayon.

 

[thesmokingman]

Lmao, all that huffing for MX4? MX4???

 

[Vya Domus]

You're not really giving us enough data, we need graphs and charts, hard numbers giving us a look at what is happening when and where. If you don't have them already, get GPU/CPU-Z and HWiNFO and post a bunch of data shots. We need to see idle and load temps for your components as well as fan speeds. I kinda agree with Vya Domus regarding your original fan temps but, idk man, two rads well in excess of 300mm and two reservoirs?? I'd expect to see lower temps more in line with thesmokingman's rig tbh.

There is definitely something odd with the GPU, I've seen people with aftermarket AIO coolers on 2080ti that get lower temps. I'd expect at most something like 55c from a custom loop like that.

 

[mrthanhnguyen]

Put liquid metal on that 2080ti. Make sure you tight all the screws all the way down.

 

[Waldorf]

@holyprof
use something like OCCT to put load on stuff (like cpu+gpu to get max psu load), also eliminates running multiple tools at once.

@John Naylor
still pushing the "set radiator as intake" huh?
multiple sites have proven it only helps lowering the loops temps by 2-3*C,
something i can easily get using LM instead of thermal paste, while having lower temps inside the case.
and once you have drives with proper controllers that pass 1GB/s, you will find lower case temps will help a lot to prevent drives from throttling,
even if the rest of the rig is under full load.
Or you want to tell me your running the heater on full in the middle of summer so that you car gets a couple degrees ;ess, you know, experts say its the best...

@Sayón
im using arctic stuff since about 2003 (and their TMPs were some of the first products), and while its better than stock/toothpaste/anti-seizure stuff (dont even try that one),
its thermal capacity is well below anything that would cause a temp drop thats larger than 5-10*C (depending on cpu/cooler/load etc).
so there was clearly something "botched" on the previous install for this to show such a significant drop.
i guarantee,. no one would spend their time/money to use liquid metals to gain <10*C, instead of using something like 5$ MX4.
completely ignoring the fact that i have yet to see any major site doing a TMP comparison/review getting results anywhere close to yours.
Liquid Metal: thermal conductivity: 79 W/mK
TP like MX4: thermal Conductivity: 8.5 W/mK
(so unless the laws of physics don't apply to your room... )


@mrthanhnguyen
not sure that is the issue, iand i wouldn't even bother to put it on the gpu,
as you would need to clean chip/block and reapply the LM at least another 1-2 times before block/chip have saturated (to prevent the LM to turn "solid" over time.

 

[xtreemchaos]

MX4 is a laugh, you can get cheap paste from china that performs better but i give mx4 is better than toothpaste by far ...

 

[HammerON]

Let's not get off topic with what TIM is better or not please. Take it to PM.

 

[petedread]

 

[thesmokingman]

I just ran realbench @8gb. Your loop looks like its doing it's job but your gpu is a definite problem. A waterblocked gpu should stay below 50c on average in my experience. Yours is way too hot. The cpu temp on the other hand looks fine, though idle seems a lil bit high for my taste. Btw, how are you monitoring temps, etc with hwinfo? Do you know what cpu sensor you are reading? Mine displays cpu die average however my loop runs off of water temp and not cpu temp.

The cpu is stock and gpu is daily overclocked.

At idle...
ambient 24c
w temp 28c
cpu 30c
gpu 28c


After Realbench
ambient 24c
w temp 33c
cpu 75c
gpu 33c

 

[Vya Domus]

still pushing the "set radiator as intake" huh?
multiple sites have proven it only helps lowering the loops temps by 2-3*C,

There is no real difference between having it as an intake or exhaust however if you have one radiator as an intake and the other as an exhaust temperatures will definitely be worse by more than 2-3 degrees.

 

[Waldorf]

In reality, pushing for 2 maybe 3c less on cpu temps will hurt other things much more.

Just tired of him always posting it regarding what was actually the question/ignoring real world use/impact.
e.g. on a gaming rig the best cooling should be on the gpu, as just passing 54*C will start to drop boost clocks,
and even IF, the gains on the cpu clocks from having the it even 5*C cooler is maybe what 50ish Mhz?
Most of the time gaining 50 MHz on the gpu will have bigger FPS impact than 100 Mhz on the cpu.

 

[EarthDog]

View attachment 154214

Why is this an image? WTH?

 

[xtreemchaos]

i still dont think your gpu temp isnt that bad, the power draw from your card must be high as there is 8+8+6 power connectors , have you found a post with someone with alike setup ? 65c is nothing for a gpu, my 1080 with on 8pin can go upto 55c on heavy vr sessions.

 

[Waldorf]

not bad?
i can hold 55C with a 120mm rad on AIO under fur load.
sure, not a ti, but it's a ftw with the beefiest power setup i've seen on xx80 cards.

 

[Vya Domus]

65c is nothing for a gpu, my 1080 with on 8pin can go upto 55c on heavy vr sessions.

My 1080 tops at 50 with a cheap AIO, that's about 200W worth of heat and a single 120 rad. This guy has as much radiator real-state as you can possiblly fit in a case and he's getting 65c on a GPU that let's say consumes at worst 100 more watts ? Yes, the CPU is also in the loop so let's say 400W in total, that's still nothing for the amount of cooling he has.

So let's get this straight, he has a system that dumps about twice the heat my 1080 does but he has what, 6-7 times the radiator area ? Come on, something is definitely wrong. You can expect 65c from a really good air cooler.

I am fairly convinced the GPU temps hes getting are because of uneven pressure of the waterblock/ bad paste job or something of the sort.

 

[xtreemchaos]

both your replys are well valid, but me i wouldnt worry too much about that temp id just enjoy the cracking pc hes built, but then again as you say Vya if the WB isnt seated right id have to have a look because 10c lower would be positive. my 1080 when doing benchmarks dont go far past 50c or just under but VR when playing FO4VR really do apply the heat guys.

 

[thesmokingman]

i still dont think your gpu temp isnt that bad, the power draw from your card must be high as there is 8+8+6 power connectors , have you found a post with someone with alike setup ? 65c is nothing for a gpu, my 1080 with on 8pin can go upto 55c on heavy vr sessions.

At 65c on the gpu, that defeats the point of investing in a 160 dollar gpu block. 50c is already already on the high side, let alone 65c which is ridiculous. You've clearly never experienced a blocked gpu in a proper loop.

lol, the dual 2080ti 3970x render monster I built last Dec which has only has 120mm more rad space than the OP's build, cools double the load of the OP and yet, both 2080ti max out under a 5 HOUR RENDER at 50c.

Build: 3970x, dual 2080ti, 8TB m.2 RAID = Render Monster

I'm building a render machine for work on a variety of applications. It will be an investment in reducing completion times in rendering and content creation workloads. This is a build for an old college buddy who works in the industry. On to the component list. Also, going to start with two...

hardforum.com

 

[xtreemchaos]

"carm down man" its just a opinion bro i dont get so hung up on temp its not like its cooking theres more to life, if it ant catching fire alls good to me i was just saying it wouldnt worry me too much . buddy i dont get blocked gpus or anywhere in my loops because i dont use crappy liqids just distilled water.

 

[petedread]

@EarthDog Oh, just seen that in the thread. I did not think I had posted that. I was writing a reply in Windows OneNote on my phone to post later when I had a signal. Not sure how it ended up being posted but it is not my finished post. I think I was going to add something about my CPU being on Creator mode everything on auto.
@Vya Domus I will be going back to all rads as intake (just for the dust filters) so there is as you say a few C gain to be had there.
@thesmokingman I did the HWinfo rtss set-up as you suggested. And I will fill that side fan/radiator spot with exhaust fans.

My delivery has been held up because something was not in stock and they did not get around to telling me until I phoned them today. So what I will do, just for the experiment to see what happens is to use the Noctua thermal paste that I have to reseat the GPU block. And then do it again when the better paste arrives. I will do it in the morning it's too late to start now.

 

[Waldorf]

@xtreemchaos
except at 65C you already dropped at least 50Mhz, and on a ti that hurts.

@petedread
depends on what your priority is.
cpu clocks? then do intake.
if not, front/side/bottom as intake (no matter what if case/rad) and top/rear as exhaust.
i had used the side as exhaust for the rad, top just case fans, but switching to case intake for side and put rad on top (exhaust) helped a bit.

 

[Nike_486DX]

I understand that you are amazed EarthDog but i'm not kidding. Otherwise I wouldn't post that.
I did 2 tests with my old AIO...Coolermaster 240 RGB with the standard paste and 1 test with the MX-4.
With the new Coolermaster ML360 RGB and MX-4 it dropped 28 degrees.
It must be the paste because all settings where the same in all tests.

whats the ambient t? it has to be the same, otherwise such measurements would be total BS.

But hey, you know what? I was using Arctic MX4 and my temps were really high, now i applied some
"Thermal Paste grey Silver Silicone Grease Compound Heatsink"
and i got 30C drop in temps, using same cooler, same fan rpm and config, even did the same paste application !!
...Well at that time i moved from Ethiopia to Norway, but that is another story


but honestly, how are you supposed to compare thermal pastes if u dont even take the air t into account?