Jump to content
Corsair Community

High Temps Using Corsair h100i pro rgb


SsheartleSs

Recommended Posts

Hello guys. I just bought a corsair h100i pro to cool my i7-8086k. I correctly installed the pump following the instruction, i applied a pea sized drop of noctua nt h1 thermal paste on cpu, wiped the pre applied cooler paste and fitted it not too togh, but not loose either

 

Installed icue, i maxed the pump rpm(water temp is 30C), maxed the fans and started stress testing a i7-8086k 5.0ghz 1.29V manual oc, MCE disabled as well as all C states disabled

 

Prime95 Small FFTPS 96C after 2 mins

Aida64 Extreme 91C after 40 min

Witcher 3, No Man's Sky, Hitman 2, Battlefield V 80C (1hour each)

 

I dialed down my oc and voltage and there was a difference of -4,5C. I feel something is very wrong with the cooler performance though at idle it stays at 36C and at full load water temp rises at only 37C

Link to comment
Share on other sites

If your coolant temp is 36C when you start and only 37C at the peak, then you have a contact issue. It should be about +6C for 100% CPU load and up to +10C when gaming if you have a lot of GPU waste heat warming the case. You probably need to take off the block and check the TIM spread.

 

Also, no more Prime until this is sorted and even then, there are better tools unless you have a specific purpose in doing a custom run. It's very easy to trip the power limits on Z370 boards or some will pile the voltage on to make sure it's stable at all costs. I don't know the GA settings as well as others, but check your actual Vcore during load in Aida compared to the 1.29v set. Most boards have an overshoot and a little is OK. 1.36v would be far too much. AIDA or Intel XTU will be a better tool for doing controlled stress tests during analysis. They also have graphical representations of the data to help pinpoint issues. AIDA's blend test is a bit chaotic by design, but Intel XTU produces a nice smooth sine wave like temp curve. It should be rolling waves. If you are seeing steep mountains on each up cycle, that again suggests contact issue.

Link to comment
Share on other sites

The coolant is 30 when i start and 37 when i finish the stress or gaming

 

No waste heat in the case, i have 3 intakss front, 1 intake bottom, radiator as exhaust, 1 fan as exhaust in the back of the case, temp in case is 22C, gpu temp is 65C under load

 

Vcore isn't overshooting, LLC is High, max voltage overshoot 1.31max, 1.28min

 

So i took out the case, the cooler, cleaned the paste(was evenly distributed by the looks of it and not too much) with isopropyl alcohol, reapplied again but this time arctic silver 5 a pea size, screwed the pump in place better and fired aida64 stress test with fpu. Max temp was 90 after 2 mins and never rised above that for an hour

 

The bracket of this cooler is a little loose on my mobo(gigabyte makes the thinner than competition)

 

For thus vcore the temps are a little off in my opinion. I pass aida test, realbench, prime95 without avx, gaming without any instability but temps seems to me a little off

 

I had prior to this a cooler master masterliquid for my past cpu i5-8600k 5ghz and it never went above 85C

Link to comment
Share on other sites

Your room temperature may be 22C. I will bet money that your case temperature is not 22C. And ... to be clear ... unless you have a liquid cooler on your GPU, you do have waste heat in the case and a LOT of it. That GPU may only be 65C under load but it's staying there because it is efficiently dumping the waste heat. It will also generate heat at idle, just not as much.

 

The bracket is a little loose on all motherboards. That's fine.

Link to comment
Share on other sites

FYI i have a thermometer inside the case, to monitor my temps and airflow changes. I see that you can't understand a thing from what i wrote. With a cooler master masterliquid 240mm inferior to this corsair, the temps are at least 8C less even in stress testing. Whatever heat waste i may have in my case, an inferior cooler cooling better it already says that there is something wrong with h100i.

 

Another thing you can't understand, the cpu temps goes from 35 idle to 80C in game in a matter of seconds another thing that it is not happening with the cooler master

 

Rather than being cocky, be a better beta tester and find that the gsync interferes with your application creating lag and stutter every time a user hovers his mouse over your app, wich renders icue totally unuseable with a gsync capable monitor. Very easy to replicate, i opened a ticket for that too

 

The bracket is a little loose on all motherboards, that's why corsair agreed to send me a new better bracket for my cooler?

Link to comment
Share on other sites

I feel something is very wrong with the cooler performance though at idle it stays at 36C and at full load water temp rises at only 37C

 

 

30 to 37 is quite different and exactly what is expected. I would have skipped over the contact issue as unlikely if you had been clear. Coolant deltas are small values and you have to be precise. The re-paste was fort naught and I would not have suggested it with a 7C delta.

 

Well, I would take a very hard look at your BIOS settings and compare them to other Gigabyte users to see if there is something you missed. I don't see it and if your peak Vcore is only 1.31v, then there isn't a clear source either. That leaves the very unsatisfying possibility you have a below average 8086K or it has a less than perfect construction (more likely). You coolant delta is 7C. Even if we turn this into 720mm radiator, the most you could reduce temps by is 7C and zero delta isn't possible with internal coolers. Now at 90C every little bit helps, but not likely to point your would like. Your CPU temp to coolant differential is in the 50C+ range. That's pretty high, but not unheard of for Coffee Lake running at 1.30v. This is why people were so quick to delid on this model and my coolant to CPU differential delidded is more like +25-30C. So unless you think the cold plate suddenly won't conduct heat...

 

You can't compare a different CPU with a different cooler to this CPU and a different cooler. However, if your games go from 35 to 80C on launch, we are back to the same point. The CPU is not conducting heat out fast enough. So if the voltage side is in check, then we are looking at the physical properties of the CPU. You can try pushing down lightly on the pump block while under load to see if you get a temp drop. That would suggest a fit issue. However, I am a doubtful because this would be evident at idle as well and these 8700/8086 CPUs never sit quietly. It would ping up to 50-60C every few seconds and likely ride in the 40s. Also, if the coolant delta is +7C, heat is being conducted into the cooler.

 

If you are convinced the cooler is to blame, put your CM cooler or anything else back on. Metal doesn't stop conducting heat at will, so if its the bracket or fitment, that should be evident. Frankly, I don't really care if you take the H100i PRO back or not and I don't have the patience today to teach all the science or manners you should have learned as a child. If you can't be polite, look for help elsewhere.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...