The Corsair User Forums  

Go Back   The Corsair User Forums > Corsair Product Discussion > Cooling

Notices

Reply
 
Thread Tools Search this Thread Rate Thread Display Modes
  #1  
Old 08-19-2016, 04:37 AM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 867881
Astral85 Reputation: 11
Default H80i GT and Corsair Link frustrating me!

I'm on a clean install of Windows 10 (1607) and I'm not getting the temps I previously had.

I'm running an i5 4690k with the H80i GT clocked at 4.2 GHz and 1.215 V.

Previously running Prime95 small FFT s I wouldn't get any higher than 70°. Currently I'm getting 77-80°

I can't think what is causing the increase in temp. The pump speed is on balanced and I never needed performance mode for the temps I was getting.

The various fan speeds in Corsair Link don't make any difference to the temps.

Has anyone got any ideas what's causing the increased temps?
Reply With Quote


  #2  
Old 08-19-2016, 04:50 AM
red-ray red-ray is offline
Banned
red-ray's PC Specs
 
Join Date: May 2014
Location: England (GMT+1)
Posts: 7,148
POST ID # = 867883
red-ray Reputation: 81
Thumbs down Not enough information

You need to provide more information, especially the coolant temperature.

Can you post a screen shot of the CL4 [home] tab?

Last edited by red-ray; 08-19-2016 at 09:22 AM. Reason: typo
Reply With Quote


  #3  
Old 08-19-2016, 08:02 AM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 867904
Astral85 Reputation: 11
Default

Sure, here's CL4. This is probably the average coolant and cpu temps when the machine has been running most of the day. So far my gaming temps are what I would expect but I'm positive my temps we're not so high in Prime95.

Reply With Quote


  #4  
Old 08-19-2016, 09:08 AM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 11,296
POST ID # = 867910
c-attack c-attack Reputation: 116
Default

You are running an extremely strenuous test that puts the majority of the load on the CPU voltage side of the system, not the cooler or it's ability to remove heat. If you want to compare cooling performance, you need to compare load water temperature to your pre-test starting water temperature (H80i GT Temp). Small FFT is always going make the voltage on die the limiting factor.

As for the change, you likely had chipset drivers available from Asus in relation to the new Windows build. People have reported a small bump in CPU temps on the new drivers, but I cannot confirm if this is true on a Z97 platform. You also need test in identical circumstance to compare CPU temperatures. Your room and case temps need to exactly the same or you need to adjust for the differences. A +3C rise in room temperature also means +3C on your cores at all levels.

There is nothing in your screenshot to suggest there is a problem with the cooler and your water temperature is pretty good for a H80i compared the other case temps. I am assuming it is rear mounted as exhaust. It is doubtful there is a problem with the cooler or you wouldn't be able to run small FFT for more than 60 seconds.
Reply With Quote


  #5  
Old 08-19-2016, 10:07 AM
HOODedAssault HOODedAssault is offline
Registered User
HOODedAssault's PC Specs
 
Join Date: Dec 2011
Location: Ontario, Canada
Posts: 1,270
POST ID # = 867927
HOODedAssault HOODedAssault Reputation: 139
Default

My guess would be you're running a different version of prime95 then you were before. Or your remembering your prime blend test temps because small fft will get about 10° hotter then blend. Also version 28.9 compared to version 27.9 can get 5° or more hotter.
Reply With Quote


  #6  
Old 08-19-2016, 07:36 PM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 867997
Astral85 Reputation: 11
Default

Quote:
Originally Posted by HOODedAssault View Post
My guess would be you're running a different version of prime95 then you were before. Or your remembering your prime blend test temps because small fft will get about 10° hotter then blend. Also version 28.9 compared to version 27.9 can get 5° or more hotter.
No, I'm still running 26.6 and no, I'm positive I was getting lower temps in the small fft tests. I'm running P95 from a backup drive but I don't think that would make any difference.

I fired up Intel burn test and hit 80° almost instantly but IBT is even more strenuous than P95 I think...

Last edited by Astral85; 08-19-2016 at 07:38 PM.
Reply With Quote


  #7  
Old 08-19-2016, 08:01 PM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 867998
Astral85 Reputation: 11
Default

Quote:
Originally Posted by c-attack View Post
You are running an extremely strenuous test that puts the majority of the load on the CPU voltage side of the system, not the cooler or it's ability to remove heat. If you want to compare cooling performance, you need to compare load water temperature to your pre-test starting water temperature (H80i GT Temp). Small FFT is always going make the voltage on die the limiting factor.

As for the change, you likely had chipset drivers available from Asus in relation to the new Windows build. People have reported a small bump in CPU temps on the new drivers, but I cannot confirm if this is true on a Z97 platform. You also need test in identical circumstance to compare CPU temperatures. Your room and case temps need to exactly the same or you need to adjust for the differences. A +3C rise in room temperature also means +3C on your cores at all levels.

There is nothing in your screenshot to suggest there is a problem with the cooler and your water temperature is pretty good for a H80i compared the other case temps. I am assuming it is rear mounted as exhaust. It is doubtful there is a problem with the cooler or you wouldn't be able to run small FFT for more than 60 seconds.
I had wondered if there might be some extra voltage occurring that I'm not aware of but all monitoring programs look to be reading the correct voltage. I had thought about an increase in room temp but we're talking about an almost 10° increase.

I checked back on an earlier troubleshooting thread of mine and I was reporting a very stable 65° under small fft at an even higher clock of 4.5Ghz.

Yes, rear mounted. Right so probably nothing hardware wise wrong with the cooler but maybe more something driver or settings related somewhere...
Reply With Quote


  #8  
Old 08-19-2016, 08:08 PM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 868000
Astral85 Reputation: 11
Default

Ok this is weird. On a clean boot this morning I run P95 small fft again with a max temp now no more than 60°. I set my OC using Asus AI Suite and I'm curious as to whether the issue lies there but I can not prove anything.
Reply With Quote


  #9  
Old 08-19-2016, 08:31 PM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 868005
Astral85 Reputation: 11
Default

Screenshot of P95 after a few passes.

Reply With Quote


  #10  
Old 08-19-2016, 10:28 PM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 11,296
POST ID # = 868033
c-attack c-attack Reputation: 116
Default

Is there a voltage reduction on this run?

Is the current OC AI suite based? Or was it the prior with higher than expected temperatures?
Reply With Quote


  #11  
Old 08-19-2016, 10:36 PM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 868034
Astral85 Reputation: 11
Default

Quote:
Originally Posted by c-attack View Post
Is there a voltage reduction on this run?

Is the current OC AI suite based? Or was it the prior with higher than expected temperatures?
No change in voltage. All OC's have been with AI suite.
Reply With Quote


  #12  
Old 08-20-2016, 12:34 PM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 11,296
POST ID # = 868080
c-attack c-attack Reputation: 116
Default

I am looking at your VID under the Intel 4690K/Voltages section. This is HWMonitor's representation of your BIOS "core voltage". It looks like like a steady and peak value of 1.10v. That likely accounts for the dramatic drop in CPU temperature compared to the 1.2x value.

I don't have enough to data to interpret the original bump in CPU temps. However, you need to careful when comparing. A shift in median or average CPU core temperatures is noteworthy, but must be viewed against any change in room or case temperatures. Peak values may not be reliable or consistent depending upon the application used and/or the CPU model. Those random peaks are the strays and not necessarily indicative of your general performance. You really need a line graph to make sense out of it.

I have a clear 3-4C bump in average core temperatures after the Win 10 AU update and corresponding Asus chipset drivers. I don't know if you have gone down that path.
Reply With Quote


  #13  
Old 08-21-2016, 09:13 AM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 868157
Astral85 Reputation: 11
Default

Quote:
Originally Posted by c-attack View Post
I am looking at your VID under the Intel 4690K/Voltages section. This is HWMonitor's representation of your BIOS "core voltage". It looks like like a steady and peak value of 1.10v. That likely accounts for the dramatic drop in CPU temperature compared to the 1.2x value.

I don't have enough to data to interpret the original bump in CPU temps. However, you need to careful when comparing. A shift in median or average CPU core temperatures is noteworthy, but must be viewed against any change in room or case temperatures. Peak values may not be reliable or consistent depending upon the application used and/or the CPU model. Those random peaks are the strays and not necessarily indicative of your general performance. You really need a line graph to make sense out of it.

I have a clear 3-4C bump in average core temperatures after the Win 10 AU update and corresponding Asus chipset drivers. I don't know if you have gone down that path.
I'm not sure why it was only using 1.10v on this occasion, it may be a display error as I usually always see my correct voltage which I have locked into Ai suite at 1.215v. I have something enabled that controls voltage based on load that could be responsible but unlikely.

I don't really think its room temp type of issue, the temp differences are too large. I'm positive that on some occasions/system boots that something is not working correctly causing the temps to skyrocket. I should note that I have to manually change the voltage in Ai suite on every reboot as it always defaults to 2.275v which is pretty high.

It's possible that although lowering the voltage in Ai suite that sometimes it hasn't actually worked and the voltage is higher than thought/reported. Just a thought, need more tests/prime95 runs.

Last edited by Astral85; 08-21-2016 at 09:17 AM.
Reply With Quote


  #14  
Old 08-21-2016, 06:17 PM
c-attack c-attack is offline
Registered User
c-attack's PC Specs
 
Join Date: Jun 2014
Posts: 11,296
POST ID # = 868206
c-attack c-attack Reputation: 116
Default

Unfortunately, this is the root of the problem with using AI Suite to alter BIOS settings. Some values, even when available in AI Suite, cannot be altered from the Windows environment. A physical shutdown of the power supply may be necessary to make the change. This is something the BIOS will do automatically when set and something AI Suite cannot, except from the tuning process and reboot.

I can't make grand sweeping statements about which values you can change. Your Z97 BIOS may behave differently than my X99 version in it's relationship with AI Suite. My general advice to people is you can use AI Suite auto-tuner to overclock. But once you do that, make changes in the BIOS only. A further complication is the odd way some BIOS settings appear when tuned with AI Suite. The values may appear skewed or non-sensical. The best thing to do is either 1) use AI suite to OC and then don't touch it. or 2) Use AI suite to OC, take notes of its settings, and then do a better job yourself. There are really only 3 or 4 values you need to touch to make a successful overclock. It's that last 5% of possibility that causes all the trouble. AI Suite won't get you there either.

What is controlling voltage based on load? That sounds highly likely to have some impact on your results.
Reply With Quote


  #15  
Old 08-21-2016, 11:15 PM
Astral85 Astral85 is offline
Registered User
Astral85's PC Specs
 
Join Date: Aug 2016
Posts: 356
POST ID # = 868225
Astral85 Reputation: 11
Default

Your observations look to be correct. For some reason when I run Prime95 the cpu voltage is deciding to only run at 1.100v despite being set at 1.215v in Ai Suite. This is where I'm getting the temps of 65°. Other times it's using the correct set voltage resulting in the higher temps. The question is, why and what is it setting random voltages? I'm starting to question the reliability of Ai Suite, a full bios overclock is generally favoured correct?

I need to check on that voltage regulator setting.

Last edited by Astral85; 08-22-2016 at 08:39 AM.
Reply With Quote


Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -4. The time now is 01:04 PM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2020, vBulletin Solutions, Inc.