Higher CPU temps - should I worry?

[honanhal]

I recently noticed that my computer seemed to be running hot during demanding P3D situations (no surprise there, right?). In checking temps with Real Temp I found that it was running around high 70s (C) in demanding scenarios with short spikes as high as 83C.

(It's a 6700k overclocked to 4.6, using a Z170 OC Formula present for 4.6 that uses 1.3V adaptive voltage -- a bit high but I've never once gotten a BSOD, and temps used to be totally fine, so I haven't had reason to complain...and cooled with a Noctua NH-D15S.)

83C seemed a bit high, and I hadn’t cleaned my computer out in a year, so I did. Not much dust inside, and after closing it back up temps were basically the same. I had a record of my RealBench results from when I set up the computer last year, so I ran it again to compare. Bottom line, it's not P3D that made the difference - my CPU is running hotter than it was a year ago.

Tried to attach screenshots, but google drive won't work, so:

2016 max temps for four physical cores with 30-min benchmark:
75   77   73    76

2017 max temps for four physical cores with 30-min benchmark:
79   81   77    82

So, 4C higher on three cores and 6C higher on the last one. A significant difference.

As to why it's happening, since no hardware has actually changed in that time, my case fans are ancient (seriously - 6 years old!), and at least one made some rattling noises recently, so I'm perfectly willing to believe that while they're still running they may not be cooling as well as they used to. At some point I'll replace them, or more likely just stick the whole computer in a fresh case. But I'm not in a hurry to do that, because that's a big project and a huge pain.

So my question is: should I be worried about these temperatures? I mean, I'll keep an eye on this and check temps from time to time, but are these temperatures going to do any damage? If I run P3D a few times a week at mostly upper half of the 70C range, and other games/programs that don't even approach these temps, is that OK for the time being? Really appreciate any advice. I build my own computers because flightsim is demanding of hardware and because I'm cheap, not because I enjoy it...

James

[ziggy]

Nope

[RichieFly]

The differences could be ambient conditions. A higher room temperature and humidity are a factor in cooling efficiency.

 

[MDFlier]

How clean are the blades on those ancient fans? They are airfoils and just like ice changes the profile of airplane wings, dust buildup changes the profile of fan blades. I've seen lots of fuzzy stuff stuck to them. If you are going to keep the system (the case in particular) for a while, just replace the lot of them. 

Just for grins, I would remove the CPU cooler and replace the thermal compound. It does lose some of it's effectiveness over time.  

[honanhal]

1 hour ago, MDFlier said:

How clean are the blades on those ancient fans

Actually, pretty clean -- I've cleaned them several times in that 6-year stretch, including just this week. They don't have fuzzies, at least, although I'm sure they've accumulated some sort of film over the years.

1 hour ago, MDFlier said:

Just for grins, I would remove the CPU cooler and replace the thermal compound. It does lose some of it's effectiveness over time.  

After a year would it have lost that much effectiveness? I don't want to reseat the cooler unnecessarily, since it's a hassle and I'm always terrified I'll screw something up, but if you think that could account for those temperature differences over a year...

[martin-w]

Yes, ambient temp may well be the difference. Always check room temp when testing and consider the delta temp.

When testing, variables should be kept the same. You say this is with P3D, so is it EXACTLY the same scenario? Same P3D version, same P3D settings, Same graphics card settings, same test scenario? 6C hotter is so little, that only a slight variable at work would do it.

May also be worth re-seating the D15 and applying new TIM.

As for a rattling case fan, doubt it would make a difference. You would be surprised how little difference a single  fan makes. 

Don't worry about brief temp spikes. This is normal. Not sustained long enough to be an issue. it's how these CPU's function these days. 82 degrees will not damage your CPU. TJ Max is about 105 degrees I recall. However, it is a bit hotter than I would like. Mid 70's is the max I prefer. Except when running a demanding stress test when I allow higher.

[MDFlier]

That's why I said "for grins". It's just sometimes one of those things that you would think would never work, works. And for the record, by "lose it's effectiveness" I wasn't thinking about it going bad, I was thinking about it from the perspective of it "moving" due to the heat from the CPU and pressure from the heat sink pressing down on it. It goes through a heating and cooling cycle every time you use the computer. I turn mine on every day. If I leave and come back, I turn it on again. That's probably close to 500 heating and cooling cycles per year. On a micro level, the expanding and contracting of the CPU and heat sink as they heat and cool might cause the compound to migrate or squeeze out. 

Besides that, sometimes just the process of checking something fixes your problem. I've been in the computer business 32 years. I've fixed a boat load of machines accidentally. I took something apart and found no issue, then put it back together and It starts working, I have no idea what was really wrong. Loose connection somewhere I guess. Point is. Try a few things, re-measure. If it is a real change, it is either due to a SW change over the last year, a physical change in the cooling system, a change in another component in your system, or an internal change in the CPU. Too bad you don't have a system image backup of the day you ran that test. It would be perfect for ruling out a change in your SW. I doubt that a film on the blades would mess with your air flow. The fuzzy stuff can. 

Edit: One last thing. Are you 100% sure that you haven't changed any obscure BIOS settings over the last year? 

[honanhal]

Well, you guys were right — it looks to have been the ambient room temperature. It’s winter and I guess the central heating was a little more aggressive than I’d realized...

I cracked a window open (just enough to let in some cold air — still totally comfortable in the room without warm clothes) and found that P3D’s momentary peaks were now down to 79C. I’ll rerun the Real Bench Test later with the window cracked again, but I fully expect the results will be very similar to my 2016 ones.

Thanks for the help, all!

James

[martin-w]

Good news James.