Bennylava

Here are some numbered questions about heat on intel chips...

1. Is it true that 65nm chips typically kick out a lot more heat than 45nm chips when overclocked?

2. What is the farthest one can safely push a 65nm Core2duo in terms of heat? Take my e4300 for example. It is a 65nm core2duo with 2Mb L2 cache. With my 24/7 OC of 3.4Ghz, load temps in prime95 blend test stay are around 68-70c. In your opinon, is that too high? I want the chip to last about 2 years.

3. What is the farthest you can safely push a 45nm Core2Duo in terms of heat? I also have a 45nm core2duo chip that runs at about 73c (maximum)in prime blend test. Is this too high if i want 2 years out of the chip? I was under the impression that the 45nm chips are actually a little bit tougher than the 65nm chips when it comes to heat. Is there any truth to this at all?

Thanks - Benny

linskingdom

1. In general, yes.
2. This needs to refer to load temp. Each stepping has its max thermal limit. E4300 is a L2 stepping. It has max thermal limit is 61c. 68-70 is high in this case.
3. Same as item 2. Which 45nm chip do you have? Is it e8400? It has C0 and E0 stepping. C0 is 72c and E0 is 74c.

__________________
Push to the limit one more time..

Teufle

65 usually runs hotter then 45.
The rest, I dont know...AMD ftw

Bennylava

Hmm. What do you think will happen if i just leave it the way it is? A shortened life span? Is it drastically shortened, or just kinda shortened? Cause those were prime95 blend test results. Games don't get CPU's quite as hot as prime does heh.

DeadSkull

45nm core 2s are built on the highK process (same as i7) so they should* be tougher temperature wise.

__________________

Q9650 Club

HD 4870x2 w/ Accelero FS - $270

transhour

i think the life expectancy on these chips are like 35 years any ways, overclocking and heat wise, might shave 15 years off their life span, so unless you plan on keeping this rig for more than 20 years, i say you should down your overclock.

__________________
There are three kinds of lies: Lies, damned lies, and benchmarks.
never go as5 to mouth
ASrock[/COLOR] Pro/Extreme/Deluxe/SuperComputer club, P55 and X58 welcome[/B]"]http://www.overclock.net/intel-motherboards/588268-asrock-pro-extreme-deluxe-supercomputer-club.html
Xeon Glory

Bennylava

You sure about that? Some people (though i guess they could just be used to way older chips) seem to think that they only last about 5 years, and OC and heat dwindle this down to 2 years or less. I was just wondering if anyone knew this stuff for certain. If it's like you say, why don't people just run their OC up as high as they want and not really worry about voltage or heat as long as they are within somewhat reasonable limits? I see a lot of people around here go up high, then get scared and back off. Not being sarcastic or anything like that, just trying to learn something new lol.

wildfire99

i dont know about 35 years but im pretty sure most common chips have a life span well over 10 years and i honestly wouldnt worry much about that being a issue, the foremost problem i see with over-overclocking is chip-degradation where your chip wears down over time and cant get as good of stable overclocks.

__________________
OCN Veteran for Revival of Chat!

Blameless

Depends on the chip and how far it's OCed. Generally, for a given clock speed 65nm chips will run hotter.

At their max OCs, 45nm chips often consume more power and put out more heat. Smaller manufacturing processes actually have more leakage.

Depends on the voltage and speed you use. Temperature, in and of it self does not say much, as it's only one of many factors that affects electromigration.

Unless you are going crazy with the voltage, 70C will likely last quite a bit longer than two years on that chip.

Again, there are other important factors.

If you are running at near stock speeds and volts, you can run the chip 24/7 nearly at it's tjmax (say 95C if you have your tjmax set to 100) and reasonably expect the CPU to last 4-6 years.

Even with a fairly substantial OC and voltage increase, 72C (28C from tjmax) isn't going to kill a chip in two years.

No, not really. Intel's lower end chips typically do not have the heatspreader soldered to the die like the higher end ones do, so they are less tolerant of heat, but there really shouldnt be a significant increase in the heat a 45nm chip can take otherwise.

This process just helps offset the leakage caused by the process shrink. I'm not convinced it leads to a CPU that can stand higher absolute temperatures.

T_1000nutss

I was wondering the same thing as Bennylava in regards to temps on the E8400 chips, I have A Xeon E3110 Which is basiclly the same chip with a new name on it. which runs at 73c - 74c on 99% load while testing it with prime 95 and real temp. Is 74c on 99% load safe?