6700k load temps question...... help

[elkido122]

I have the 6700k cpu with a kraiken x61 setup in a nzxt h440 case. my load temps get to about a max of 60c does this seem normal for stock settings and no overclock. idle is around 22-27c in a room temp of around 72F

 

[BonzaiDuck]

I'm still becoming familiar with my 6700K rig, but that does sound about right, depending on what load you're putting on the system.

The Skylake is manufactured following the pattern that arose with Ivy Bridge followed by Haswell: no more use of Indium solder for the consumer chips between the heat-spreader and the processor die. Skylake uses a polymer thermal interface material.

By contrast, you can order a processor from Silicon Lottery online and pay $50 to have it de-lidded and then re-lidded with Liquid Ultra. It will buy you between 8C and 18C improvement in temperatures. At the same clock-speed and voltage, my chip is about 5C cooler than a retail-box CPU without the re-lid in a test-bed featuring an external dual-fan EXOS water cooler. I'm using a ThermalRight LG Macho heatpipe cooler.

I'm not suggesting that you do this, although Silly-Lotts will perform the $50 service if you mail them your chip. Otherwise, those are the temperatures you can probably expect under stock load conditions, and they're going to increase if you overclock.

What stress test are you using to generate those temperatures? That will give us a better idea of your expectations or an assessment of your temperature results. I can revise my comment with that information.

 

[elkido122]

thats just under load while playing bf1 which we all know currently puts alot on the cpu, also i monitored it using my kraiken cam software and hwmonitor.

 

[[DHT]Osiris]

60's probably fine. I have an OC'd 6700k, I forget what my load temps were last time i checked it, but I have a large watercooler block and the fans curves up quite a bit under substantial load, which would indicate that it does indeed get warm 'enough'. You're 40% under cutoff so I wouldn't sweat it too much.

 

[elkido122]

i just didnt know if maybe i need to think of a better case or reapplying the thermal paste. my radiator in this case is top mounted with the fans pushing into the radiator.

 

[crashtech]

60°C is nothing for a modern CPU, it can do that and more for years on end. You are totally fine until and unless it starts threatening to bump into 100°C under the most extreme conditions.

 

[elkido122]

well 60c and its not even summer time, what happens when summer hits and the room temp is higher. ?

 

[SirCanealot]

CPU temp will go up mostly in line with room temp. So unless your room is going to be 30 degrees hotter I wouldn't worry too much

 

[Yuriman]

A 40c increase is ~72F so if your room temperature increases 72 degrees F then yes, you might need to be concerned about your CPU's temperature.

 

[EXCellR8]

check voltage in BIOS, might be able to knock it down a few notches if the board is sending too much juice.

 

[nerp]

60c is perfectly fine. You could probably lower temps a little but it would more to bring comfort to you, not the CPU.

 

[elkido122]

What's the best method for applying thermal paste to a 6700k

 

[Yuriman]

My opinion - grain of rice size/shape in the middle of the IHS, oriented the same way the core is. The heatsink will spread the paste out when you install it.

Stolen from google:

 

[crashtech]

A round ball, 1/4" or 6mm in diameter is just about equally effective as the "rice grain" method, imo. Ideally you will have a tiny bit of excess TIM squeeze out of the sides of the IHS, but the TIM won't quite make it to the corners. Maximizing contact from the IHS to the HSF/water block without creating an excessive mess is the goal. One might think pre-speading the TIM might be a good way, but this only traps air bubbles.

 

[BonzaiDuck]

A round ball, 1/4" or 6mm in diameter is just about equally effective as the "rice grain" method, imo. Ideally you will have a tiny bit of excess TIM squeeze out of the sides of the IHS, but the TIM won't quite make it to the corners. Maximizing contact from the IHS to the HSF/water block without creating an excessive mess is the goal. One might think pre-speading the TIM might be a good way, but this only traps air bubbles.

I still stick with ICD diamond paste. Their recommendation avoids spreading the unwieldy stuff, applying a pea-sized blob in the center of the IHS. Their idea about this is that spreading may create air pockets, as you describe. I confess I never gave up using an expired credit card to apply an even coat to both surfaces, and then I add something close to a pea-blob before I seat the heatsink base. If I failed to follow their prescriptions to a T, I don't think I've ever missed out on good contact and effective heat transfer.

The stuff oozes from between IHS and HSF base to form a bead, which develops the consistency of that gummy caulk people use to seal their bathtubs to the vinyl or tile flooring of the bathroom. I may even waste it, but it's a minimal investment in something that can actually be rejuvenated with common silicon or oil-based TIM, and re-used.

I've fallen into an irresistible habit of lapping my processor-caps to bare copper, so all the minor ill-effects of the diamond abrasive -- making the small-print identifiers unreadable -- doesn't matter.

Silicon Lottery actually sold me a CPU with bare-copper IHS after my consultation and their willingness to provide one.

 

[crashtech]

@BonzaiDuck , everyone has their favorite way of doing things, and some habits are hard to break, but I think you might find that skipping the credit card procedure will save time and material without any discernible change in thermal conductivity. A back to back test might be fun just for your own edification, and perhaps ours as well if you decide to test it.