Building a new PC

Yep, that was my latest attempt to fix the issue. It helped a little at idle, but the fan kicks into high gear pretty often due to the temperature spikes. I'll probably just have to deal with it until either:

  • The fan dies on its own due to excessive use.
  • It annoys my wife and I enough that it's worth it to get a quieter, cooler fan.
  • Intel releases a fix to the processor (many have requested a fix be provided, though I have no idea how they would fix a hardware temperature issue other than to intentionally cap the processor usage/speed).
 
Doing a little more research and more measuring, this is the best cooling I can get for this system. Unfortunately, Gigabyte has some housing right where a bigger fan would need to go, so the bigger fans/pipes/fins probably would not fit on this motherboard. I don't think it's a deal breaker, though. I added a second fan to the fins, and the temperatures dropped another 5 C during normal use. They still get to 70 C under 100% load, but that's not that high for this CPU. Other reviews I've read have said that the lids used on these CPUs are connected to the case in a manner that limits thermal dissipation, so without delidding the chip, I probably won't see much better temperatures.

I'll just have to bear the noise from the fans from time to time or make sure music/tv/movies/etc. are on in the background.
 
A good, cheap solution would be to replace the stock with a Noctua NF-F12. I was unhappy with the noise of my Corsair H60, but swapping out the fan for an NF-F12 made it dead silent.

I'd recommend looking at an AIO, they're normally quieter and perform better. Plus they have small CPU blocks that won't have any issue fitting on around annoying blocks on the mobo
 
Tekno Venus said:
A good, cheap solution would be to replace the stock with a Noctua NF-F12. I was unhappy with the noise of my Corsair H60, but swapping out the fan for an NF-F12 made it dead silent.

I'd recommend looking at an AIO, they're normally quieter and perform better. Plus they have small CPU blocks that won't have any issue fitting on around annoying blocks on the mobo
Click to expand...

I'll keep the Noctua NF-F12 in mind. I'm not sure how it would mount to the fins, though. The current fans have clips on them.

I avoid AIO. I've seen too many display cards killed by a leak.
 
Tekno Venus said:
Easy :p Cooler Master Hyper 212 EVO + Noctua NF-F12 PWM - YouTube

Really? I've never seen any issues with AIOs, especially not nowadays. They're a really well established market nowadays.
Click to expand...

It was one of the first AIOs that I witnessed firsthand, and I saw several forum threads with similar experiences from other users around the same time. They may be better now, but I'd still lean toward air cooling over water cooling as being safer.

If these fans end up being replaced, I only have one set of brackets and I would like to continue with the two fan setup. Based on your suggestion, I am leaning toward Amazon.com: Noctua L-Type Premium Quiet CPU Cooler_ Retail Cooling NH-L9x65: Electronics which has high reviews including one 7700k user who said temps were 55 C at 100% load. It has the added bonus that it's a top down cooler, so less space issues, and it will probably do a better job of keeping the motherboard components around the CPU cooler.
 
I have very good airflow through the system with two 120mm pulling air in across the board and one 92mm pushing warmer air back out.
Click to expand...
It does not matter how much cool air you jam into the case, if you don't exhaust it out, heat builds up. I would move one of those 120mm fans to the back to increase pulling heated air out and see how your temps do (which I agree, are not bad - but I would like those highs to come down a bit too).

Did you apply as thin a layer of TIM (thermal interface material) as possible? It is important to note the best transfer of heat occurs with direct metal to metal contact. The purpose of TIM is to only fill the microscopic pits and valleys in the mating surfaces to push any trapped insulating air out. Any excess TIM is actually in the way and counterproductive to heat transfer efficiency. Also, depending on the TIM formulation, some TIMs require a curing period where over a few days of heat-up/cool-down cycles, temps can come down by as much as 4 - 5°.

You also stated some concern over how fast the temps go from cool to peak highs. I'm not surprised. It only takes a few clock cycles for CPU temps to go from cool to over heated. And since that i7-7700k cruises long at 4.2 billion cycles per second, a lot can happen in 1 second. And because fans are mechanical devices, it takes time for even the best fans to compensate - which they don't begin to do until the BIOS senses, then directs the fan speed to ramp up. So rapid temperature increases can be expected - but temperature drops should be nearly as quick as well.

I agree that AiO coolers are much better than they used to be, but anything with hoses can develop leaks, either at the fittings or even the hoses can be punctured, flawed, or over time become brittle and crack.

I personally don't like water cooling solutions - though my position is probably biased based on past history and experience with clients, and not any recent problems.

1. Too often after new builds with alternative cooling, users are extremely diligent at performing regular inspections to make sure there are no leaks and everything is working fine. The problem is, after a year or two of no problems found, those regular inspections fall by the wayside right when components start to age and inspections should increase! We saw many older systems that developed leaks, some causing severe (and scary) damage, some where the coolant just evaporated, and even a couple where really nasty mold took over. :(

2. Designers intentionally cluster heat sensitive and heat generating devices around the processor socket where they can take advantage of the air turbulence provided by the processor fan. If the user does not properly address case cooling to provide an adequate flow of cool air through the case, those devices may not receive sufficient cooling. This can result in instability or increased aging. Back in the day when electrolytic caps were commonly used, this may have resulted in leaky or bulging capacitors at a higher rate than those boards using conventional/OEM cooling.​
 
Digerati said:
Any excess TIM is actually in the way and counterproductive to heat transfer efficiency.
Click to expand...
I watched an interesting video about TIM a while back, the results from their experiments were very interesting. Obviously it's just one experiment, but I was surprised at the results.


Basically... thermal paste application technique isn't quite as important as you think
Click to expand...
 
Digerati said:
I have very good airflow through the system with two 120mm pulling air in across the board and one 92mm pushing warmer air back out.
Click to expand...
It does not matter how much cool air you jam into the case, if you don't exhaust it out, heat builds up. I would move one of those 120mm fans to the back to increase pulling heated air out and see how your temps do (which I agree, are not bad - but I would like those highs to come down a bit too).
Click to expand...

I could probably put a couple fans on top or on the side of the case to vent heat out. The case is pretty well ventilated even with only one exhaust fan. It has plenty of exhaust vents, and I feel a good airflow of warm air coming out the top/sides of it from the CPU fans.

Digerati said:
Did you apply as thin a layer of TIM (thermal interface material) as possible? It is important to note the best transfer of heat occurs with direct metal to metal contact. The purpose of TIM is to only fill the microscopic pits and valleys in the mating surfaces to push any trapped insulating air out. Any excess TIM is actually in the way and counterproductive to heat transfer efficiency. Also, depending on the TIM formulation, some TIMs require a curing period where over a few days of heat-up/cool-down cycles, temps can come down by as much as 4 - 5°.
Click to expand...
I put a pea sized amount on and smoothed it to an extremely thin layer using a credit card to collect excess while keeping it evenly distributed.

Maybe overkill based on Stephen's post, but that's how I was taught. :grin1:

Digerati said:
You also stated some concern over how fast the temps go from cool to peak highs. I'm not surprised. It only takes a few clock cycles for CPU temps to go from cool to over heated. And since that i7-7700k cruises long at 4.2 billion cycles per second, a lot can happen in 1 second. And because fans are mechanical devices, it takes time for even the best fans to compensate - which they don't begin to do until the BIOS senses, then directs the fan speed to ramp up. So rapid temperature increases can be expected - but temperature drops should be nearly as quick as well.
Click to expand...
Yes, it goes from 25 C to 50 C quickly and almost as quickly back down to 25 C.

Digerati said:
I agree that AiO coolers are much better than they used to be, but anything with hoses can develop leaks, either at the fittings or even the hoses can be punctured, flawed, or over time become brittle and crack.

I personally don't like water cooling solutions - though my position is probably biased based on past history and experience with clients, and not any recent problems.

1. Too often after new builds with alternative cooling, users are extremely diligent at performing regular inspections to make sure there are no leaks and everything is working fine. The problem is, after a year or two of no problems found, those regular inspections fall by the wayside right when components start to age and inspections should increase! We saw many older systems that developed leaks, some causing severe (and scary) damage, some where the coolant just evaporated, and even a couple where really nasty mold took over. :(

2. Designers intentionally cluster heat sensitive and heat generating devices around the processor socket where they can take advantage of the air turbulence provided by the processor fan. If the user does not properly address case cooling to provide an adequate flow of cool air through the case, those devices may not receive sufficient cooling. This can result in instability or increased aging. Back in the day when electrolytic caps were commonly used, this may have resulted in leaky or bulging capacitors at a higher rate than those boards using conventional/OEM cooling.​
Click to expand...

Yep, that's why I'm considering getting a top down cooling solution. My motherboard temperatures don't seem too high (most temperatures are in the 30s; there is one around 70-75 C when the processor is in use, which is probably a sensor near the processor).
 
Thanks for that video. The results were interesting, for sure. Too bad they did not state which TIM they were using.
 
Did he say that, or did you recognize the tube? I note too there are a couple different IC Diamond formulas too.

Hmmm, Malwarebytes keeps yelling at me when I visit the innovationcooling.com website.:confused2:
 
He didn't say, but that's the thermal paste of choice on that channel and has been for years. I recognised the tube he was using.
 
Last edited:

Has Sysnative Forums helped you? Please consider donating to help us support the site!

Back
Top