Temperatures - Don't worry so much - Page 5 - Overclock.net

gamer11200 · 01-14-09

Quote:

Originally Posted by kimosabi

A chip is not made to last for only a year or so. This "a year" term works for an OCer but not for the average Joe.

Yeah, and most of the time, the chip that we buy costs >$200 so we don't feel like killing it. I prefer to still have the chip in working condition so then I can sell it to get more money for a new chip and make someone happy in the process

CyberBoy · 01-14-09

Quote:

Originally Posted by XFreeRollerX

+1

but if you get random shutdowns while loading, that would also be considered throttling

My chip licks 74*C once in a while at 4.4Ghz, did the same at 4.2Ghz on air....

Chip is running 11 months strong so far

74c on a E8400. Jesus man. I never broke 55c at 4.4ghz

MaXiMiUS · 01-14-09

So.. I shouldn't worry about my temps so much and just go to town with my CPU overclocks? CPU is at 3.3GHz in this screenshot.. 2.66GHz stock.

Explaination of the screenshot:
CPU is at 100% load (after several hours of cpuburn)
in0: Detected CPU voltage (1.28125V in BIOS)
fan1: CPU fan speed
temp1: Case temp
temp2: Heatsink temp?
Core 0: self-explanatory
Core 1: self-explanatory

zooterboy · 01-14-09

Quote:

Originally Posted by kimosabi

Too much heat makes it harder for the CPU to run stable. A chip @ 4gHz and 50C works more efficiently than a 4gHz chip @ 75C. It's all a bunch of transistors in there and they want the coolest environment they can get. Especially when overclocking past manufacturer's "safe" specs.
How do you people figure that they can get such high OC's using LN2, DICE, Phase etc? Could it possibly be because they have lower temps? And even then they kill their chips quite often.
A chip is not made to last for only a year or so. This "a year" term works for an OCer but not for the average Joe.

Anybody that's been around this site for a while has seen this debate more than once. What's the most dangerous is voltage, not heat. If you combine the two, the effects are amplified obviously, but with 45nm chips you'll typically reach max voltage way before you reach max temps.

I believe what you consider safe is entirely dependent on you. If you worry about what temp is too high, you probably shouldn't even be approaching the lowest considered temp limit. Sure, we can all argue over who is right and who is wrong, but the fact remains that if you are overly concerned about cooking your chip, you have no business overclocking it. It's the risk we all take as soon as we go over the stock settings; if I couldn't afford to cook my CPU and replace it, it wouldn't be overclocked.

zooterboy · 01-14-09

Quote:

Originally Posted by CyberBoy

74c on a E8400. Jesus man. I never broke 55c at 4.4ghz

But are you in Arizona?

The temps on the hood of my truck are nearly that in April. And it's not even overclocked. much.

Crooksy · 01-14-09

Quote:

Originally Posted by Havegooda

Sarcasm much?

~Gooda~

Why do you say "much" on the end, its so annoying. Talk properly.

trueg50 · 01-14-09

Very true.

Temps on the x3220 Xeon chips (Q6600 with a Xeon name) is a max of 85C, so regular 65nm Core 2's should be around there.

As for voltages, it really does very, though the 65nm Core 2's are tanks. Many have run at 1.6 volts for extended periods, and there is one case of some one running at 1.7 volts for 6 -12 months and it was still working fine.

**Retrospekt** · 01-14-09

Quote:

Originally Posted by gamer11200

I disagree. I had my P4 prescott running with a stock cooler. Playing a simple game like UT2004 caused this thing to top 80 celsius. After a few months, the motherboard got fried.

A hot processor has nothing in correlation with your motherboard frying.

facepalm.

megajesus · 01-14-09

Voltage > heat.

oneluvballer21 · 01-14-09

A few people have hit the right areas already... at stock voltage you can probably not worry getting your temps up to 90c, and higher you should be careful. But we're overclockers, and we generally don't like stock voltage. If you raise the voltage, your max temp will lower, being the point where your system gets unstable. Raise the voltage more, the max temp lowers more. Why can't we take a CPU and pump 1.8v into it for a sweet OC? Probably because you aren't running LN2. If you want higher-than-stable voltages, you need to significantly lower your temps. You want crazy 1.8v voltage?? You need sub-zero temps. They go hand in hand. You can't just say watch your temps if you are at a stock or below-stock voltage, cuz unless you are running without a cooler, you probably won't hit high enough temps to worry too much... your max CPU OC will likely be voltage limited before you get to those temps, assuming an average/stock cooler. But you raise the voltage and that max temp area gets lower, just like your CPU's load temps get higher. Once you've maxed your stable OC/voltage, you may well be in the range of 65-70c as a max safe temp... depending on the quality of your chip, of course. Higher quality silicon can take get higher clocks on less voltage, so the max temp may be skewed a touch from average chips.

Anyway, temps matter as much as voltage because they are correlated when it comes to system stability at a max OC. To push a CPU past your max, you'll need better cooling to stabilize the voltage you can give the proc.

		Overclock.net - Overclocking.net > Cooling > Air Cooling
Temperatures - Don't worry so much

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)