• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

How cold is too cold?

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

jack_ruby

New Member
Joined
Mar 19, 2002
A couple questions i have seen everything from watercooling to liquid nitrogen used to cool proccessors.

1. so at what temp will i start to do damage to my CPU?
2. Also What CPU and Mobo combo are best for Overclocking?

Thanks
Jack_Ruby
 

kevin_bouchard

Member
Joined
Sep 5, 2001
Location
manitoba
chances are anything under -10 will damage your chip or shorten the life, I would worry more about temps rising and falling. If you can get a temperature that is pretty stable you should be fine.
 
OP
J

jack_ruby

New Member
Joined
Mar 19, 2002
Kevin,
Thank you for your input. I'm going to sound like the noob I am, is that -10 degrees Fareinhight or -10 Degrees Celsius?

Thanks Again for your help.
Jack_Ruby


Anybody else have any input on this?
 

Bender

Mysteriously Changing Senior
Joined
Dec 19, 2000
Location
In Thelemac's Basement Eating the Chickens
jack_ruby said:
Kevin,
Thank you for your input. I'm going to sound like the noob I am, is that -10 degrees Fareinhight or -10 Degrees Celsius?

Thanks Again for your help.
Jack_Ruby


Anybody else have any input on this?

I assume he means Celsius. I had my celeron [email protected] idling at -21c with a peltier and it was absolutely fine. As long as you don't have huge temperature fluctuations you should be ok even at very low temps. I'm hoping to have my Athlon XP running below zero by tomorrow. I just got a 226 watt peltier and a 50 amp astron power supply. Wish me luck. Its scarry sealing up the socket area for fear of condensation.
 
OP
J

jack_ruby

New Member
Joined
Mar 19, 2002
Bender said:


I assume he means Celsius. I had my celeron [email protected] idling at -21c with a peltier and it was absolutely fine. As long as you don't have huge temperature fluctuations you should be ok even at very low temps. I'm hoping to have my Athlon XP running below zero by tomorrow. I just got a 226 watt peltier and a 50 amp astron power supply. Wish me luck. Its scarry sealing up the socket area for fear of condensation.

Well best of luck to you. and thanks for the input ill have to search for some threads on how to seal the socket.

Later
Jack_Ruby
 

Staz

Member
Joined
Mar 16, 2001
Location
Lafayette, La.
kevin_bouchard said:
chances are anything under -10 will damage your chip or shorten the life, I would worry more about temps rising and falling. If you can get a temperature that is pretty stable you should be fine.

Kyroteck and Vapochill both set thier systems to run at -40.

BTW -40 is the same F or C
 

JoeCrappa

Member
Joined
Jun 11, 2001
Location
Union City, CA
Thats weird. i never knew that -40 was the same for F and C...i guess you learn something new everyday. i was about to write a post contradicting that. but then i did the calculations and you're totally right!
 

ButcherUK

Member
Joined
Nov 8, 2001
Location
York, UK
Chris-the dude said:
-276.16 C is too low.:cool:

Not only too low, but also impossible... -273.16C is absolute zero.
Minimum temp you should aim for for prolonged running (i.e not just a day etc.) is -50C. You can go lower for short periods with LN2 etc. if you're careful.
 

Tiger

Member
Joined
Apr 6, 2001
Location
UK
Staz said:


Kyroteck and Vapochill both set thier systems to run at -40.

BTW -40 is the same F or C
Sorry this is not correct. The conversion formula is C/5=(F-32)/9.
-40C = -62F
 

Tiger

Member
Joined
Apr 6, 2001
Location
UK
Someone in Sweden tried this on a P4 a month ago. They cooled to -200C for a period of a few hours and the proc failed. There is some conjecture that this has to do with differences in contraction ratios for the different materials used to make the proc. It is generally accepted and has been done practically by someone in Japan that a T-Bird can be run at -150C. The machine was run for a few days at this temp. The ideal temp is around -100C. There is not much further benefit below this as the curve of the speed vs temp starts to rise vertically at about -150C and therefore no matter how much colder it is no further speed is going to be coaxed out of the proc.
 

Brant

Member
Joined
Jan 27, 2002
Tiger said:

Sorry this is not correct. The conversion formula is C/5=(F-32)/9.
-40C = -62F

Oh, it is definitly correct. -40 C and -40 F are the same.

F=(9/5)C+32 C=(5/9)(F-32) same as C/5=(F-32)/9 so check your math -40 C= -40F
 

Brant

Member
Joined
Jan 27, 2002
Once you get to liquid nitrogen temps you start to get super condutors(perfect conductor). Now if they only made ceramic super conductor processors.
 

tft

Member
Joined
Feb 22, 2002
Location
R.I.
Huh. It's a good thing I read this post. I was planning on using a freezer to supercool my system.
How much life would it take off the chip if you were to run it at -20 C or something like that. Would you even notice the shortened lifespan, or would it be more like you only have a few months left after you get to the low temp.
 
OP
J

jack_ruby

New Member
Joined
Mar 19, 2002
Hey all thanks for your input. Keep it coming the more information i have on this the better. and the more of an educated decision i will be able to make. If i was able to maintain lets say around -21c
and keep it there maybe fluctuating 5 degree variation would that be acceptable or is 5 degrees to much?
thanks again
Jack_Ruby
 
Last edited:

Tecumseh

Member
Joined
Feb 26, 2002
Location
Ohio
Tiger said:

Sorry this is not correct. The conversion formula is C/5=(F-32)/9.
-40C = -62F

Sorry, Tiger. -40C = -40F exactly. Physicists haver been
using this as a handy reference point for years.
 

Tiger

Member
Joined
Apr 6, 2001
Location
UK
Brant said:


Oh, it is definitly correct. -40 C and -40 F are the same.

F=(9/5)C+32 C=(5/9)(F-32) same as C/5=(F-32)/9 so check your math -40 C= -40F
Yes sorry you are correct. It so happens that -40C=-40F but not all -C temps = -F temps
 

Tiger

Member
Joined
Apr 6, 2001
Location
UK
tft said:
Huh. It's a good thing I read this post. I was planning on using a freezer to supercool my system.
How much life would it take off the chip if you were to run it at -20 C or something like that. Would you even notice the shortened lifespan, or would it be more like you only have a few months left after you get to the low temp.
It doesn't shorten the lifespan it actually lengthens the lifespan because electromigration is reduced.
 

ssjwizard

Has slightly less legible writing than Thideras
Joined
Mar 12, 2002
well ive ben told that anything below -40C the silicon paths begin to freeze to a point where it will degrade cpu performance that doesnt nessicarily mean that its breaking the cpu it just means that the cpu isnt functioning as well as it could. id say about -30C is probably a good temp with a variance of 9 or so should be fine. id recomend using a supercooled liquid though since the temprature fluxuation will be less. or even a cold liquid with and a peltier with a low delta T like if you can keep a 300W load at -10C grab a 200W pelt with a Delta T of 20C and an athlon and watch your temps stick around -30C all the time