Jump to content
Corsair Community

Hydro H60 V2 expected core temp under stress?


Polyfun

Recommended Posts

When idle my core temps are between 25-50 c. When I run the CPU: OCCT stress test they climb to > 80 c after about a minute. Is this to be expected?

 

Cooler fan is plugged into CPU_FAN (on max speed, about 1500 RPM) and cooler pump is in CPU_OPT (on max speed, about 4500 RPM). Cooler fan is set to intake, and I have external top and front fans to blow out. I recently replaced the thermal paste too. Ambient temp is about 18 c.

Link to comment
Share on other sites

OK, then default settings with the usual motherboard enhanced 4 cores at the turbo boost speed. 1.22 is decent auto voltage for 4790K. It might be possible to shave a few hundredths off, but I am not sure that is worth the time and hassle at this point.

 

Yes, large data set is fine. I was curious if you had used small knowing the results would drop 10C going to large. Regardless, presumably you did not put this together to run stress tests. How is your real world performance? Are there any strange behaviors like super-reactive core temps, really high spikes when idle, etc.?

Link to comment
Share on other sites

You are right on the money asking about real world experience, so here is some more context. I have been using this rig fine for the last three years for gaming. I started playing Total War Warhammer recently (CPU intensive) and that is when I started getting CLOCK_WATCHDOG_TIMEOUT BSODs. So that led to me try OCCT for the first time, which led to the questions about core temperatures, as they seem to be high under OCCT stress. Sometimes using OCCT I would also get CLOCK_WATCHDOG_TIMEOUT BSODs, so I concluded that the high temperature was causing the BSOD.

 

I now think this conclusion was wrong, and the BSOD was due to something else, e.g., voltage or XMP (which was switched on). I have now reset my BIOS to the "optimised" defaults, and XMP is off. I have also turned CPU_FAN and CPU_OPT to the max (1500 RPM and 4500 RPM).

 

I no longer get the BSOD on Warhammer, and core temp when playing is < 50 c; so OCCT is not really representative of the real world.

 

One thing puzzles me. My memory (Corsair Vengeance Pro Red 16GB (2x8GB) DDR3 PC3-19200C11 2400MHz Dual Channel Memory Kit) is rated at 11-13-13-31, but the BIOS defaults loaded it is set to 9-9-9-24. Should I worry about that? I tried to change it to 11-13-13-31 but it caused a BSOD.

Link to comment
Share on other sites

It can be a real pain to try and trace the BSOD's back to a source, but I do not think they were thermal shutdowns. The BIOS will often give you a secondary warning when that is the case. Additionally, as you approach the higher temp tier, the processor would start to throttle, the game would chug, and there would be some visceral feel things were amiss -- not the sudden blue screen out of nowhere.

 

Before I even got to the last paragraph, I was wondering about memory. MMO engines in particular seem to be sensitive to these settings. I usually know I have screwed up or gone to far when one of those crashes. 9-9-9-24 is the JEDEC standard timings for DDR3 at 1333MHz. All DDR3 memory modules will default to those values and 1.50v when you stick them in. Those are the rules. The XMP is a preset overclock for your modules. That is the 11-13-13-31 2400MHz@1.65. Nearly all motherboards support XMP and it is simply (usually) a matter of toggling the switch in your BIOS. However, the more complicated aspect is enabling XMP can also change other values like DRAM voltage, VCCSA, base clock or strap, and other things determined by the board manufacturer. Some of these are good and necessary. Sometimes one of the values is too much or at least arbitrary. Unfortunately, you will need someone with specific Gigabyte Z97 platform experience to guide you on the values or intricacies. These things are very motherboard dependent. 2400 is a good overclock for DDR3, but it is possible it is just a little too much or needs some manual fine tuning. The real question of the moment is whether you were getting BSODs when set to 9-9-9 and your frequency was the default 1333. If the BSODs were occurring at those settings, we are barking up the wrong tree and overclocking the memory is a separate, unrelated issue.

Link to comment
Share on other sites

No I was not getting BSOD with 9-9-9 and 1333.

 

I can successfully run OCCT with XMP but with Intel Turbo Boost off (so CPU is at 4.0 GHz and not 4.4 GHz). When I use XMP and 4.4 GHz I get the CLOCK_WATCHDOG_TIMEOUT. I am not sure what this means...if anything. The Totalwar benchmark only shows 1 FPS difference between 4.0 GHz + XMP and 4.4GHz + no XMP. It would be nice to get 4.4GHz + XMP stable though.

Link to comment
Share on other sites

It sounds like you are just a touch light on Vcore. Haswell CPUs have the memory controller tied into the main CPU voltage. Theoretically, a CPU running 4.4GHz and 2400MHz will require slightly more Vcore than the same CPU when running 4.4 and a lower memory frequency. People complain about the auto voltage running to high for normal use, but this is one reason why the table is generous. Try a small bump to Vcore, maybe 1.24 or 1.25v. If you were running the DRAM at 1600-1866 you likely would not need a bump, but 2400 is getting on the high side where tweaking is often required. Like always, this is not an exact science and may require a little fiddling to find the right voltage.
Link to comment
Share on other sites

Archived

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

×
×
  • Create New...