H60 Safe Temps?

Guest · Sep 24, 2014

Hi, I'm just wondering if my CPU temps using a H60 are safe. My CPU is a Intel Core i5 4690k @4.4ghz/1.18V. I ran Intel Burn Test on Maximum for 10 cycles and recieved no failures but temps topped out at 83c.

Is this ok? If not how should I proceed?

Many thanks.

Farapon · Sep 24, 2014

Totally fine

Guest · Sep 24, 2014

Seriously? Would you go as far as to say this was good for a h60?

JackNaylorPE · Sep 24, 2014

That's higher than I permit on my builds (I'm at 74C @ 4.6 Hz, 46 cache, 2400 mem, 1.385v) but I wouldn't worry too much unless it's seeing that for long sessions. Try RoG Real Bench .... it uses real life applications rather than synthetic tests.

The H55 provides a little less cooling than the Hyper 212 so don't expect a lot.

snowctrl · Sep 24, 2014

The h60 (Corsair H60?) is maybe a little underpowered for your OC, but the temps you are getting are about what you would expect - they are high, but you're unlikely to see that in every day use, gaming etc, unless your doing video encoding, 3D rendering etc

Guest · Sep 24, 2014

Yes, it's the corsair h60 to be more specific. If I wanted to reduce these temps to at least under 80c what would you suggest? At 4.3ghz and 1.6v it ran intel burn test on max for 5 cycles and topped out at 81c. Why is this? As such I'm unsure as to what will decrease the temps while maintaining 4.3+ghz.

Thanks for the help.

snowctrl · Sep 24, 2014

1.6v? I hope that was a type, and that your vCore is the 1.18v you mentioned above... Presuming you've applied your thermal paste well, there's not a lot you can do... Will it take an additional fan? And what is the rest of your case cooling like?

Do consider though, that IntelBurnTest et al are not typical usage scenarios... why not leave it and monitor it in day to day use, see how it does

Guest · Sep 24, 2014

Yeah, it was

. Meant 1.16. Yeah, I guess I'll just monitor it then, I've read a lot about problems with the h60 and was hoping I'd fitted it correctly. I Guess this is just its limit. Thanks for the help.

snowctrl · Sep 24, 2014

No worries, best of luck with it

JackNaylorPE · Sep 24, 2014

A good air cooler would do it.

Phanteks PH-TC14-PE, Noctua DH-14 or 15, Cryorig R1.

A 240/280 mm AIO would do it ... avoid the hi rpm ones, look at the H110 .... the Swiftech H220-X beats everything,m is quiet and can be expanded to include GFX and other water blocks.

h2o79 · Sep 24, 2014

If you add one more fan to your h60 so you get push\pull, it should give you a few degrees. It did for me anyway.

Davil · Sep 25, 2014

Only other thing you can do is de-lid as shown in the video below. God speed.

https://www.youtube.com/watch?v=sqfpgUK9Ldo&list=UU0_swmlWWBAJofDN1nOUcjA

Guest · Sep 25, 2014

Thanks for the tips guys. Idling at around 8-12% (resource monitor) CPU usage I'm getting 32c (20c ambient), up to 40% usage can see me hit the low 40s; playing Crysis 3 for around 45 mins my max temp was 72c. Bear in mind also that at idle my 2 front intake fans are running at almost max speed (1000rpm) and my cpu fan is running at 1100ish/2000 rpm. Taking into account that as my board does not offer adaptive voltage, i'm still running at 1.18v at idle while my cpu frequency does fluctuate. Is this anything to worry about?

Many thanks

Davil · Sep 25, 2014

Are the front two fans connected directly to the power supply? If so that's why they are always at max speed. If not you can change their settings in the BIOS or by using a fan controller. I would honestly be very suprised if your motherboard didn't have at least an offset setting, but no 1.18 isn't a big deal. 1.4 is the limit most people go to for daily use so you should be just fine. Voltage doesn't necessarily mean heat either, really it's the loading of the CPU that causes heat and under a full load that's the voltage you should be aiming for anyway.

Guest · Sep 25, 2014

Thanks for the answer, i've left the motherboard to control fan speeds its self. The only component I have at max rpm is the pump. I can use an offset voltage, will have to look into it more. I was wondering how the voltage effected temps at idle but you kinda answered the one. How are the temps though? Might try and go for lower clock and voltage but surely the h60 should provide better cooling no? I've seen posts where people are using 212 evos to cool i4670ks at 4.5ghz and 1.3v and getting less than 70 in games. It just seems strange to me.

Davil · Sep 25, 2014

They likely have gone through the de-lidding process I linked before. Very very very few 3rd or 4th gen i5 and i7's have decent contact under the IHS so you'll even notice looking at real temp that you probably get differences of like 5 or more degrees from your hottest to your coolest core which isn't supposed to be the case. On 1st and 2nd generation intels they used fluxless solder that was awesome and all the cores were within 1 or 2c from each other and you could use a decent closed loop water cooler and get great temperatures, not with the newer ones though. You also have to figure in ambient temperature, water cooling and air cooling both rely on ambient temperature to cool things.

The difference between the ambient temperature and the temperature of thing you're cooling is called delta, at idle my room temperature is a comfortable 25c and and my cpu is at about 30c which is a 5c delta. If my room temperature goes up to 40c though my cpu will now idle at 45c. Same concept with what you're seeing from other people too. They may be in a very well air conditioned room or a cold climate.

JackNaylorPE · Sep 25, 2014

120/140mm water coolers just don't have enough Rad to compete with the better air coolers.... see below the Hyper 212 budget cooler compared with an H55

Davil · Sep 25, 2014

I will agree the hyper 212 evo is a better option over an h60 no matter how you slice it. It's quieter, cheaper, and more effective. However that still won't solve the problem with the poor thermal paste inside of the IHS.

JackNaylorPE · Sep 25, 2014

Replacing the TIM was very common with IB, much less so w/ Haswell.... and gone with DC. With a good cooler, in bty far the majority of instances, you will run into voltage issues long before you hit any thermal limits.

I am running a 4770k with 46 CPU multiplier, 46 cache ratio, 2400 RAM at 1.385v VID....drop the cache and voltages drop significantly. When ACX instructions are present, CPU voltage can climb to 1.488. I have an EK Naked Ivy delid mount sitting unused in a box because I have no heat issues but I do have voltage concerns. The 74C max core temp under max load stress testing causes no concern ..... but the 1.4 volts does. Since delidding won't reduce voltages I have had no cause to delid and use the Naked Ivy mount I bought.

Davil · Sep 26, 2014

I must disagree again, I've found that the most common reason for Haswells to lock up is temperatures in excess of 75c. Also Haswell is actually considered to be worse than Ivybridge as far as the thermal issues go. The thermal and voltage limits should go hand in hand also, increasing your voltage is going to increase your temperatures and once you get past that magical 75c mark crashes are far more likely. I think the max rated temperature by intel is 72.72c also.

JackNaylorPE · Sep 26, 2014

Not supported by data. Max temperature is 100C BTW ...starts to throttle at 98 / 99. Might take a while but if ya read this you will see that that is not the case

http://www.overclock.net/t/1411077/haswell-overclocking-guide-with-statistics

I have slowed down and even turned off my rad fans and taken temps up to 90C+ .... passed 8 hours of ETI and 2 hours of RoG Real Bench. My ssytem has 6 temperature monitoring points and have spreadsheets full of voltage / temp data.

Lotsa peeps running high 80s on Hyper 212's

Davil · Sep 27, 2014

http://ark.intel.com/products/75123/ Tcase max temp is 72.72 throttling occurs at about 105c

We have seen very different results then. I've overclocked at least 8 of these now on separate builds and motherboards and I've found that, particularly at high OC's 4.8+, 75c or so is about when they start to have problems and throw bsods at me.

JackNaylorPE · Sep 27, 2014

I'll say.... Asus indicated they tested over 1000 Haswell's and < 10%were able to hit 4.8C so if you have multiple units hitting 4.8 outta a sampling of 8, you either got really, really lucky of I wanna meet your buddy at Intel.

I have read every one of those posts in that OCN thread and no one has reported such an issue, they tease me about setting a 75C / 1.4 v limit when they runnimng at 1.5 and 85 - 95 C all the time..... out of all the participants and the OCN group only 11% of those reporting results made 4.8 Ghz ...and of course the guys at 4.2 ain't rushing to get their names ion the list.

https://docs.google.com/spreadsheet/ccc?key=0AjXDCk5eCp1gdEdENjlDYWl6ZnV4OVlNc0lMU1V3c1E&usp=sharing#gid=0

http://www.overclock.net/t/1398276/haswell-ocing-max-temp-limits

Mine goes up to 95c and never throttles. This is during prime95. Using HWinfo64 and never seen the minclock change. The chip likes 95c no matter what air cooler I used.
Using noctua NH-D14. Room temp 72F, if I take the room temp down to 68. I can get the chip down to 93c. Running at 4400mhz 1.268v, 2100mhz ram

Now mine runs at 85 under load.. but that's IBT load. I don't think anything gets your chip hotter =P That's at 1.5v and 4.9core/4.7uncore. I don't mind because apart from synthetic load, it runs very cool. Its great for gaug'ing ultimate stability but ive ran every chip I have ever owned between 1.4 and 1.5 and have yet to have one go on me.

Davil · Sep 27, 2014

Well I have delidded every single one of them which has dropped the temperatures by about 15 degrees on average so maybe that has something to do with it. I don't really put much stock in the threads you're reading because from my experience and actually quite a few threads I've seen that keeping temperature low is the key to getting a good overclock. Obviously voltage is required too and you get to a point where you need way more voltage than I'm comfortable with to get one more tick on the multiplier, but temperature is the biggest limitation. That's why people do crazy stuff like use liquid nitrogen in contests.

snowctrl · Sep 28, 2014

Davil :

Keeping temperature low is necessary for a good OC - not quite the same as being key to it - the best way to think of it is that you want to remove all the factors that might stop a good overclock, so you really get the best from your processor, and cooling is one of those factors

H60 Safe Temps?

Guest

Guest

Honorable

Guest

Guest

Titan

Distinguished

Guest

Guest

Distinguished

Guest

Guest

Distinguished

Titan

Reputable

Distinguished

Guest

Guest

Distinguished

Guest

Guest

Distinguished

Titan

Distinguished

Titan

Distinguished

Titan

Distinguished

Titan

Distinguished

Distinguished

Share this page