Acceptable temperatures above ambient for overclocked CPU?

martinlest

martinlest

martinlest

Associate
Joined
30 Aug 2011
Posts
3
Location
EGLF
I overclocked my i7-4790K today. I went with a fairly modest 4.5Ghz, using a voltage of 1.20V. So far, things seem good (I use the PC to run X-Plane 11, almost exclusively).

As expected, the CPU temperatures reported at idle are now higher than before. At the moment it is pretty warm in London and the room with my PC in it is even warmer (2 PCs running in a small space). I am wondering though how many degrees C above ambient would be about average for a i7-4790K before and after overclocking. My idle at the moment is low to mid 40s, but the PC has been on all afternoon (and I am not sure of the temperature of my room as I write this, probably around 30 degs C).

With some tweaking of X-Plane settings, the increase in the smoothness of performance in the sim after the overclocking is very noticeable indeed, better than I could ever have hoped for, so I would not want to go back to 4GHz. I can't help but be aware of the temperature increase though. When X-Plane 11 is running, the temperatures I see in Core Temp are again about 10 degs higher on average than before I overclocked - now about 65-70C. Seems fairly safe though (no?). At 100% CPU usage, when X-Plane loads, there are short peaks of up to high 80s C, but they only last a few seconds and the temperatures go back down again. Before overclocking, those peaks were some 10-15 degrees lower however.

So does this all sound within the bounds of what would be expected (bearing in mind the higher than normal room temperature)?

Oh, the PC cooler is a Noctua NH-L12, by the way. Seems to work pretty well.. It's also (another important factor) a mini-form PC, with a ThermalTake mini ITX case and an Asus H97I-PLUS M/B.. Not so much room for air to move around of course. No HDDs however, only SSDs.

Thanks for any input.

Martin
 
Last edited:
While temperature delta is a useful metric, I'd be tempted to say absolute temperatures are more relevant when talking about lifespan/full time overclock.

70 is not a bad peak temperature when gaming, it might be nice to get that down. Temperature and noise are closely related - if you have a fan curve based on temperature then in summer, rather than getting hotter the PC will probably get louder.

Potentially you could tune your overclock, your cooling/noise and your temperatures to the ideal balance in the worst-case environment i.e. in summer temperatures.

Lowering voltage will lower temperatures but I'm not so familiar with the 4790k's range.
 
As already said, temps seem okay. CPU temp is not only dependent oh heat it generates but also dependent on air temp going into cooler, and these small pancake coolers with downflow fans often run much hotter than if fan is mounted to pull air up through cooler. Reason is airflow going down into cooler hits motherboard turning out then hits RAM, GPU etc. turning up around cooler and fan where fan pulls it back into cooler. Here is visual example:

http://www.overclock.net/content/type/61/id/2464987/width/1100/height/2200/flags/LL

** Do Not Hotlink **


I've found just turning the fan over usually lowers CPU temp by 5-10. If you want to check case airflow temp into cooler you could make something like is at end of case airflow guide in this link. https://forums.overclockers.co.uk/t...-i-put-my-temp-sensor.18564223/#post-26159770
 
Last edited by a moderator:
FT05 is one of the few cases that can support (almost) passive cooling. I wonder if putting a open ended box same size vent on top of FT05 to create a taller chamber to act as chimney so the warmed air rising would have more of the chimney effect and pull resulting in more cool air in the bottom flowing up through the 'chimney'..

I've always wanted to test CR-95 against normal tower coolers without fans in a case like FT05. I'm pretty sure they would out cool CR-95. And coolers like Le Grand Macho (Thermalright) and smaller HR-02/Macho models as well. Only place CR-95 might beat conventional towers is in cases with motherboard in horizontal orientation .. but then case fans would definitely need to be in play negating the advantage of having passive cooler.
 
I believe another member here has a Grand Macho and it does outperform the CR-95.

BTW I only put the fans back in because I was having a tidy-up and that was the best place for them. They stay off almost all the time.
 
Quartz said:
I believe another member here has a Grand Macho and it does outperform the CR-95.

BTW I only put the fans back in because I was having a tidy-up and that was the best place for them. They stay off almost all the time.
Click to expand...
Yeah, as long as you aren't working it extremely hard and/or overclocking CR-95 will do fine. Seems CR-95 is rated something around 95-100 watt TDP and Le Grand Macho (no fan) is rated 140w TDP (think smaller Macho 0/fanless HR-02 are about 115w TDP). Not sure if that is both horizontal & vertical orientation or just vertical, but I think both.

CR-95 does is unique look.
 
Hi. The PC is a mini IDX format, but the ThermalTake case and the Noctua cooler do a great job under 'difficult' conditions. I find when gaming (that is, X-Plane 11 exclusively) that Core Temp settles into showing between high 50s and very low 70s, with the occasional peak (just a few seconds) up to 80. The ambient temperature is a bit higher than normal at the moment, so those figures could be a fraction lower in cooler weather. No room really for more or bigger fans in this small case.

I have set the clock to 4.5GHz (I don't really need any higher) and 1.20V. I see some folks ahve a stable 4.5GHz o/c with a slightly lower voltage - 1.18, but from what I have read 1.20 is pretty good. I haven't yet had any issues with this setting, though a small reduction in average temperatures would be good. Would a reduction to 1.18V do anything about that? I feel I should be pleased with 1.20V, so don't want to push it.
 
No room really for more or bigger fans in this small case.
Click to expand...

Wrong! I managed to get two more case fans installed in today, so hopefully, when I get round to testing, that might make a degree (or two) of difference
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom