Corsair H80i GT AIO vs Haswell-E OC.

S

sblantipodi

2[H]4U
Joined
Aug 29, 2010
Messages
3,765
Hi, is this temperature normal for an H80i GT AIO running on a
5930K @4.2GHz, 1.280VCORE, 1.92VCCIN LLC7, 1.230V cache @ 3.8GHz.

This is after 15 minutes of realbench with case open just to exclude the GPUs hot.

h80i.JPG


As you can see the maximum CPU temp was 79°c but the hottest core was 86°C.
Water was 50°c as you can see from corsair link.

Is it normal that water was only 50°c?

Is this AIO so crappy or I have some problems?

EDIT: My old H80i GT had big problems with pump, it finally died some days ago (broken pump), I bought a new H80i GT and temp now is greater by 20°C.
With the OC I explaned and the system in signature I get an awesome temp of 70°C MAX on the hottest core after 15 minutes of Asus Realbench
I'm amazed by this small AIO.
 
Last edited:
50C Liquid temperature? that's insanely hot.. are you running the pump at what speed? what fans are you using?. you want to keep the liquid temperature at the lowest possible.. an insane increase in temps like that just mean 2 things, or the fans can't keep the radiator cool enough to dissipate the heat from the liquid or the pump its too weak that isn't able to proper flow the liquid... I've said to you couple of times, the H80 is an insufficient unit for that power hog chip. you have to go with h100i and above coolers to keep temps under more control..
 
Araxie said:
50C Liquid temperature? that's insanely hot.. are you running the pump at what speed? what fans are you using?. you want to keep the liquid temperature at the lowest possible.. an insane increase in temps like that just mean 2 things, or the fans can't keep the radiator cool enough to dissipate the heat from the liquid or the pump its too weak that isn't able to proper flow the liquid... I've said to you couple of times, the H80 is an insufficient unit for that power hog chip. you have to go with h100i and above coolers to keep temps under more control..
Click to expand...

As far as I can see from reviews h100i should give me 5 degree less at Max, not a big deal.
I'm using 2x120sp from corsair.
 
Most of the reviews out there are made with socket 115X in mind, we all know that those are tiny chips with low heat generation and insanely lower power production than an OC'd Haswell-E, but not only that, unless delid is used most of the heat is trapped inside socket 115X chips which make coolers perform with little difference from one category to other its with this kind of powerhog chips where 240mm/280mm radiator shine over 120mm radiator (even if are a thick one) and again not only that, the higher surface dissipation area from Socket 2011/2011v3 chips, play a greater role with higher dissipation area AIO..

h100i its actually (hardly understandable why).. one of the top performer AIO coolers in the market probably the best according most reviews, but we can find others with the same cooling performance at way lower noise like the swiftech H240-X which its basically a pre-asembled expandable custom loop.. in both of my socket 2011 machines the difference its pretty big from an H80i to an h110i which its the cooler i used to upgrade. not only in peak temps but also in averages and typical load.. I know your case isn't able to put any of those coolers but the truth is, if your cooler its reaching 50C liquid temp its because it doesn't have the enough cooling ability. you know, you are adding like 180W-200W+ to a 120mm radiator...
 
Couple things, even with the side off your case, unless you have something like a desk fan blowing in its not going to be as cool as an open bench with two video cards in there. Even if they aren't being taxed too hard they will still generate a significant amount of ambient heat. With 6 cores running this thing is going to be need cool air from outside the case, so is your airflow coming from outside to inside with some filtering? Also reviews are very different from site to site. This one is showing temps not to much differently from yours with a 4 core 4770k at a 4.2 overclock. Every chip is going to overclock differently, so yours may just run a bit hotter.

Also I really, REALLY like the look of your case.... but its not designed to be pumping enough air through for an overclocked 6 core and two video cards. Lian Li used to make a 3 slot 5.25 bay unit that added another fan and fit right into their cases (not sure if they still make it). It would change the looks a bit but would increase your airflow a lot. It would help get your ambient down inside the case, not sure how much it would help with your actual CPU temps though.
 
Direfox said:
Couple things, even with the side off your case, unless you have something like a desk fan blowing in its not going to be as cool as an open bench with two video cards in there. Even if they aren't being taxed too hard they will still generate a significant amount of ambient heat. With 6 cores running this thing is going to be need cool air from outside the case, so is your airflow coming from outside to inside with some filtering? Also reviews are very different from site to site. This one is showing temps not to much differently from yours with a 4 core 4770k at a 4.2 overclock. Every chip is going to overclock differently, so yours may just run a bit hotter.
Click to expand...

That test used AIDA FPU only for maximum heat, if I do the same my CPU throttle at 100°c.

Also I really, REALLY like the look of your case.... but its not designed to be pumping enough air through for an overclocked 6 core and two video cards. Lian Li used to make a 3 slot 5.25 bay unit that added another fan and fit right into their cases (not sure if they still make it). It would change the looks a bit but would increase your airflow a lot. It would help get your ambient down inside the case, not sure how much it would help with your actual CPU temps though.
Click to expand...

I'm sure that is not a case problem since I had similar temp before upgrading to Ti with a GTX980 SLI reference that blows hot air outside the case.
When I upgraded to Ti open air cards I added new fans.

now I have the accessory you are talking about:

yx05BG6sHF48H1_Et2MbMbPQr80bFdYDdIjGIBFqyiA=w1920-h1200-rw


with a SP120 in it.

To sum up I have
1x frontal Corsair SP120 that blows fresh air between the upper cards and the AIO rad
1x frontal Corsair SP140 that blows fresh air in the bottom of the case
1x upper Corsair SP140 that blows hot air outside the case
2x Corsair SP120 in push pull on the rad that blows out.

My case temp are very good. It's not a case temp problem since I have very bad temp on OCCT too with cold cards.
 
Do a quick test, and this is gonna really tell you if the unit is insufficient or a case airflow isn't good, revert your H80i Fans to intake (I hope you aren't using SP120 Quiet because that will be the main problem) and keep open the case side panel.. and monitor the Liquid temperature. if it keep going so high, then the AIO unit is the problem..
 
Araxie said:
Do a quick test, and this is gonna really tell you if the unit is insufficient or a case airflow isn't good, revert your H80i Fans to intake (I hope you aren't using SP120 Quiet because that will be the main problem) and keep open the case side panel.. and monitor the Liquid temperature. if it keep going so high, then the AIO unit is the problem..
Click to expand...

I'm using the SP120 from the old plain H80 (2800RPM) that is more powerful than the one bundled with H80i GT (2400RPM) that is more powerfull than the one sold separately by corsair (2100RPM).

I have done the trial, link says 49°c.
Hey guys, this AIO is a real crap.
Can't understand why reviews all over the net shine with temperature not greate 70c even with decent overclock.
 
There are a lot of variables between a review and real world usage, thermal paste applied, method of thermal paste used, room temperature, case airflow, case location and other couple of etc.. only room temp have a gargantuan impact in review as most use sub 15C room temp, [H] is one of the few sites that use 25C room temp..... the cooler is good, but is a no match for a monster like haswell-E and the fact that the liquid temp is so high is a good proof of that situation. In your foots I would just go with 280mm AIO and you will notice how different things will behave.. you can go with EK predator240 or swiftech h240-X which are basically custom loop pre-assembled.. in the AIO segment NZXT x61, corsair h110i... if you don't mind noise the almighty h100i vanilla (not gtx just the original h100i) will also do a fantastic job.. but my choices would be with the CLC EK or swiftech..
 
Most reviews talks about delta temp with ambient temp so ambient temp is not a factor.
I use artic mx4 and I know how to apply the paste :)
 
Delta temp it's the easiest way to fool people and toy with the people ignorance. They want to people think that a cooler delta will be always the same at 25C than 9C room temp, the lower room temp the higher cooling performance a cooler can achieve, this is specially important in water cooling as the lower the liquid temperature the better it can transfer the heat from the origin to the dissipation area...

Would you really think that with the same cooler with a liquid temp of 30C have the same potential of cooling with the liquid at 15C?.. why you think custom loop coolers perform so good? They try to keep the water as close as possible of the room temp.. best of the best custom loops are even able to maintain the liquid at room temp..
 
Last edited:
You started a thread on October 6th about this same issue. In that thread I am pretty sure the consensus was that your case and cooler combo were barely capable of the task of cooling your 5930K at the speeds you are trying to get.

http://hardforum.com/showthread.php?t=1877703

There is no magic bullet to fix the issue you are having, it's called physics. Not enough cooling area (radiator size) for the heat dump you are producing.

A good rule of thumb that someone here passed on was you need to have a 120mm radiator for every 120 watts of heat dump, for best results. You are producing around 360 watts with your cpu at that voltage and speed. A 360 mm radiator would do wonders for your setup,

As we know from the other thread you won't get rid of your premium case so your only option to get temps down will be to lower voltage and speed.
 
If I switch to an ugly and cheap Corsair 450d case with a Corsair H110i GTX aio, will I lower my temp of 15°c?
 
sblantipodi said:
If I switch to an ugly and cheap Corsair 450d case with a Corsair H110i GTX aio, will I lower my temp of 15°c?
Click to expand...

Just lower your overlock and voltage. You've brought up how ugly and non "premium" other cases are so many times that you just aren't going to be happy with a non aluminum case.

You are having a bad experience because you are using an undersized AIO cooler for that CPU, you have too much heat dumped into your case by the GPUs, and your case prevents you from getting better cooling. You can't have everything, you need to compromise somewhere. You don't fully understand what you are doing and it's borderline ignorant that you refer to the AIO cooler you have as crappy. Your decisions could be considered crappy.
 
Last edited:
sblantipodi said:
If I switch to an ugly and cheap Corsair 450d case with a Corsair H110i GTX aio, will I lower my temp of 15°c?
Click to expand...

I know isn't "premium" enough for your tastes, but with a Corsair 500R and the dedicated GPU cooling fan in the HDD bays paired with the H110i GTX you will do it.
 
Ocellaris said:
Just lower your overlock and voltage. You've brought up how ugly and non "premium" other cases are so many times that you just aren't going to be happy with a non aluminum case.

You are having a bad experience because you are using an undersized AIO cooler for that CPU, you have too much heat dumped into your case by the GPUs, and your case prevents you from getting better cooling. You can't have everything, you need to compromise somewhere. You don't fully understand what you are doing and it's borderline ignorant that you refer to the AIO cooler you have as crappy. Your decisions could be considered crappy.
Click to expand...

My ignorance is based on all the non sense spreaded up by reviewers.
There are tons of reviews that shows 300w tests on h80i GT scoring only 5°c worse than. H100i gtx and having a maximum of 30°c deviation from room temp.
I read that misleading articles and I thought that a small aio like this could do the job for a mild oc like mine

Reviewers show 30°c deviations on 300w test, I have up to 70°c deviations from room temp. More than the double of what advertised on the reviews.
 
While H80 is part of the general AIO group, it is really a LCLC (Low Cost Liquid Cooler). That is what Asetek called them in the beginning. Peeps the world over who wanted "H2O / water" cooling jumped up to buy these "water wannabee" coolers. Yes, they are "water coolers" but they are very low performance. With cheap radiators with high fpi and cheap high airflow (noisy) fans, pumps that only marginally move enough water but make noise, combined with reviewer who don't have the slightest idea how to do accurate testing and raving about how great they were many fell for the mythy. But in reality if nice custom loops are new Ferraris', CLCs are 10 year old Hugos'. Good air coolers do a better job at lower noise levels 99 out of 100 times.

I don't know what your case airflow is like or how you have this H80 (cough, choke) setup,so here is a basic guild to optimizing case airflow.
http://hardforum.com/showpost.php?p=1040832403&postcount=5
 
Last edited:
doyll said:
While H80 is part of the general AIO group, it is really a LCLC (Low Cost Liquid Cooler). That is what Asetek called them in the beginning. Peeps the world over who wanted "H2O / water" cooling jumped up to buy these "water wannabee" coolers. Yes, they are "water coolers" but they are very low performance. With cheap radiators with high fpi and cheap high airflow (noisy) fans, pumps that only marginally move enough water but make noise, combined with reviewer who don't have the slightest idea how to do accurate testing and raving about how great they were many fell for the mythy. But in reality if nice custom loops are new Ferraris', CLCs are 10 year old Hugos'. Good air coolers do a better job at lower noise levels 99 out of 100 times.

I don't know what your case airflow is like or how you have this H80 (cough, choke) setup,so here is a basic guild to optimizing case airflow.
http://hardforum.com/showpost.php?p=1040832403&postcount=5
Click to expand...

my case airflow is now very good but I agree that the H80i GT is a terrible AIO and that most reviewers does not know how to test this things.
I don't know if they do the test with cold water, if I do the test when I have started the PC since few minutes with cold water, my maximum temp is 10°C less,
is after 30 minutes that the real bad performance is shown.
 
Last edited:
sblantipodi said:
my case airflow is now very good but I agree that the H80i GT is a terrible AIO and that most reviewers does not know how to test this things.
I don't know if they do the test with cold water, if I do the test when I have started the PC since few minutes with cold water, my maximum temp is 10°C less,
is after 30 minutes that the real bad performance is shown.
Click to expand...

I see a lot of sites doing AIO testing on open test benches, not enclosed cases. Try this... Take the side panel off your case and point a desk fan in there so it is blasting the whole are with air. Run your test again. What are your temps? The difference between that temperature and the temperature with your side panel on is the penalty from your case.
 
Reviewers generally test in a case calling it "real world". Well, the only "real world" involved is their own. They use room ambient as baseline air temperature instead of the actual air temperature going into cooler or radiator This is liking at bedroom at thermometer to see how warm the kitchen is. Might be the same, but probably not. Which is the reason I use a temperature sensor directly in front of cooler to monitor it's intake air temp. If CLC is mounted to draw air from outside of case it has room temperature air, but if dumping heated air into case. If CLC is mounted as an exhasut it is using heated case air but usually also increasing case airflow with it's fan/s.

The end result of their testing is now a comparison of coolers, but a comparison of how their system (their real world) performs with different coolers.

Now, if they controlled cooler intake air temp to be +/-1c, then calculated delta it would be reasonably accurate to compare coolers.

Speaking of which, have you actually monitored the air temp going into H80 and GPU?

If it take 30 minutes to get hot, my money is on case airflow is not good enough and case is heating up. I don't think CLCs don't have enough water volume to take 30 minutes of heavy load to stabilize.
 
Ocellaris said:
I see a lot of sites doing AIO testing on open test benches, not enclosed cases. Try this... Take the side panel off your case and point a desk fan in there so it is blasting the whole are with air. Run your test again. What are your temps? The difference between that temperature and the temperature with your side panel on is the penalty from your case.
Click to expand...

I tried putting the aio ouside the case, temp does not improve if not by a 1 or 2°C.
 
doyll said:
Speaking of which, have you actually monitored the air temp going into H80 and GPU?
If it take 30 minutes to get hot, my money is on case airflow is not good enough and case is heating up. I don't think CLCs don't have enough water volume to take 30 minutes of heavy load to stabilize.
Click to expand...

As I saied I tried the AIO outside the case like an open bench, temp does not improve.
I don't agree with you, all water loop require time to get the water hotter and to see the real performance.
 
Get a good air cooler. X99 Deluxe has 75+mm center CPU to near side of PCIe socket. Only problem is your 7FNWX only has 160mm CPU clearance, so unless your RAM is 25mm or less a 140mm fan will not fit over it .. and Vengeance LPX is 35mm +3mm for socket =38mm. 120mm fan will be tight fit.
SST HE01 is 160mm, has no front fan so will fit.. NH-D14 will fit. Alpenfohn K2 / Deepcool Assassin and Assassin II will fit.

Where are you located.
 
*The*Dark*Lord* said:
put the rad outside the case and test the temps then..
Click to expand...

done it, temp improves by 1°c or 2°c.

doyll said:
Get a good air cooler. X99 Deluxe has 75+mm center CPU to near side of PCIe socket. Only problem is your 7FNWX only has 160mm CPU clearance, so unless your RAM is 25mm or less a 140mm fan will not fit over it .. and Vengeance LPX is 35mm +3mm for socket =38mm. 120mm fan will be tight fit.
SST HE01 is 160mm, has no front fan so will fit.. NH-D14 will fit. Alpenfohn K2 / Deepcool Assassin and Assassin II will fit.

Where are you located.
Click to expand...

I really don't like huge air cooler :(
 
sblantipodi said:
done it, temp improves by 1°c or 2°c.



I really don't like huge air cooler :(
Click to expand...

You need to pick something from the following list:

  • Lower the overclock and voltage
  • Buy a cooler with a larger rad and a case to fit it in
  • Live with the high temps
  • Try reapplying the thermal compound
  • Convince Corsair the cooler is defective and have them send you another one to try so you can have the same issue
 
Last edited:
Ocellaris said:
You need to pick something from the following list:

  • Lower the overclock and voltage
  • Buy a cooler with a larger rad and a case to fit it in
  • Live with the high temps
  • Try reapplying the thermal compound
  • Convince Corsair the cooler is defective and have them send you another one to try so you can have the same issue
Click to expand...

I think that I will live with high temp.
While gaming with a fast speed fan profile I never exceed 70c, just stress testing is unmanageable.
 
thread updated.

My old H80i GT had big problems with pump, it finally died some days ago (broken pump), I bought a new H80i GT and temp now is greater by 20°C.
With the OC I explaned and the system in signature I get an awesome temp of 70°C MAX on the hottest core after 15 minutes of Asus Realbench
I'm amazed by this small AIO.
 
Much of that 20c change is result of new pump moving more coolant. ;) The longer a pump runs, the more the impellers wear, the lower it's flow rate becomes. And with these CLC pumps moving barely enough coolant even when new, as flow slows down heat goes up.
 
doyll said:
Much of that 20c change is result of new pump moving more coolant. ;) The longer a pump runs, the more the impellers wear, the lower it's flow rate becomes. And with these CLC pumps moving barely enough coolant even when new, as flow slows down heat goes up.
Click to expand...

I don't this that this is my case, my pump runned for too few to see some sort of wear on the impellers.
A plastic impeller can't wear that fast, I used it very very few.
 
sblantipodi said:
I don't this that this is my case, my pump runned for too few to see some sort of wear on the impellers.
A plastic impeller can't wear that fast, I used it very very few.
Click to expand...
They can wear very fast, especially if air gets into the pump and impeller and housing heat up causing them to melt. ;)

There has been very little real change in these CLC pumps H80 and H80i and H80i GTX are almost identical. There has been more change in the looks on the outside then on the inside. ;)
 
doyll said:
They can wear very fast, especially if air gets into the pump and impeller and housing heat up causing them to melt. ;)

There has been very little real change in these CLC pumps H80 and H80i and H80i GTX are almost identical. There has been more change in the looks on the outside then on the inside. ;)
Click to expand...

I had the problem with an H80i GT, that H80i GT runned for very few months with bad temperature since new, now I bought a new HH80i GT since the pump on the old unit died.

Now I have 20°C, never had this excellent temperature with the old unit, neither when it was new and I tried everything.
Reseating, changing paste, try to moving the cooler to try to move air bubbles away from the cooler and so on.
 
Did you RMA the old one? Sounds like it was defective.

But then I think all CLCs are defective. :D A top tier air cooler is better, quieter, has less problems, and cost less than a CLC 999 times out of 1000. Well, maybe only 998 times out of 1000. ;)
 
doyll said:
Did you RMA the old one? Sounds like it was defective.

But then I think all CLCs are defective. :D A top tier air cooler is better, quieter, has less problems, and cost less than a CLC 999 times out of 1000. Well, maybe only 998 times out of 1000. ;)
Click to expand...

Bullshit. No air cooler could offer this performance, at least not without putting a hige heatsink on the CPU that voids any cool look with windowed case.

I don't RMA the old one since I had no time for downtime.
 
sblantipodi said:
Bullshit. No air cooler could offer this performance, at least not without putting a hige heatsink on the CPU that voids any cool look with windowed case.

I don't RMA the old one since I had no time for downtime.
Click to expand...
RMA it anyway. You can sell it to someone else and recoop some of your loose.

No bull pucky involved. A simple mid-level cooler like TRUE Spirit 140 Rev.A or even a Matterhorn Pure will perform as well as H80i. I know, many reviews show CLCs giving better CPU temps. But there are very few reviews that do even remotely accurate testing. For example they use room ambient as baseline air temp, not the actual cooler / radiator intake air temp. Even on open bench the room air temp is usually a few degrees cooler than cooler / radiator intake. In a case it is like cooking in the kitchen and going into closed bedroom to look at thermometer to see how warm it is in the kitchen.

As for "cool" looks, to me CLCs look like the junk they are. But that is personal likes and dislikes. You like CLCs, I don't.

Important thing is you are happy with what you have.

Oh! if your new H80 fails, I've got one setting in the closet that hasn't been used sense I finished testing it. You can have it for the cost of packaging and postage.
 
doyll said:
RMA it anyway. You can sell it to someone else and recoop some of your loose.

No bull pucky involved. A simple mid-level cooler like TRUE Spirit 140 Rev.A or even a Matterhorn Pure will perform as well as H80i. I know, many reviews show CLCs giving better CPU temps. But there are very few reviews that do even remotely accurate testing. For example they use room ambient as baseline air temp, not the actual cooler / radiator intake air temp. Even on open bench the room air temp is usually a few degrees cooler than cooler / radiator intake. In a case it is like cooking in the kitchen and going into closed bedroom to look at thermometer to see how warm it is in the kitchen.

As for "cool" looks, to me CLCs look like the junk they are. But that is personal likes and dislikes. You like CLCs, I don't.

Important thing is you are happy with what you have.

Oh! if your new H80 fails, I've got one setting in the closet that hasn't been used sense I finished testing it. You can have it for the cost of packaging and postage.
Click to expand...

thank you for the offer but as I saied I don't need other H80i GT since I can RMA the broken one too.
I don't know why you don't like H80i GT, we don't need to argue on taste, everyone has his personal taste, the only fact is that 70°C on the hottest core on an overclocked 5930K with 1.280V and 1.960V VCCIN, 1.250V cache is simply amazing.
 
sblantipodi said:
Bullshit. No air cooler could offer this performance, at least not without putting a hige heatsink on the CPU that voids any cool look with windowed case.

I don't RMA the old one since I had no time for downtime.
Click to expand...

I agree...these AIOs perform much better than a huge air cooled HSF. I recently put a small 92mm AIO on an i7-6700k (on a system I upgraded to give away to family for Christmas) and it idles at 20 degC and is anywhere from 45-50degC under load (stock clocks), using the stock TIM that came on the AIO. Keep in mind that this is for the smallest AIO there is at only 92mm.

Pretty damn good if you ask me...;)
 
lutjens said:
I agree...these AIOs perform much better than a huge air cooled HSF. I recently put a small 92mm AIO on an i7-6700k (on a system I upgraded to give away to family for Christmas) and it idles at 20 degC and is anywhere from 45-50degC under load (stock clocks), using the stock TIM that came on the AIO. Keep in mind that this is for the smallest AIO there is at only 92mm.

Pretty damn good if you ask me...;)
Click to expand...

I know that there is no real reason to do it, but on my new H80i GT I removed the default thermal pad and applied the Noctua NT-H1.
 
You must log in or register to reply here.
Back
Top