CPU temp mean anything performance wise....

[Slash621]

besides not melting..... does a processor at 35c run faster than an equivelantly clocked processor at 45c.....

I'm guessing that colder is faster (why else would we need watercooling).

 

[Burninator]

sounds like a good question
I would like to know also
*bump*

 

[dguy6789]

no. Run at 60 c would be as fast as 20c, as long as it is stable, any temp below 65 is fine imo. Make that 60 because mainbaords are sometimes a bit off on the sensor. I would rather run a pc at 60c quiet then 30 c loud.

 

[fredtam]

A cpu will run the same speed as long as it doesn't get "too" hot and become unstable. People use water cooling to provide more of a window before the cpu becomes unstable due to heat. If your 2.8c for example is running 60c at full load with stock cooling and you overclock it will reach an unstable temperature before a water cooled chip that runs 50c under full load. The water cooled chip will be able to overclock more before temps became an issue. However most chips will hit a speed wall long before the temp becomes an issue. So I guess the other advantage to water cooling would be that the components run cooler at all times which may prolong the life of components but most components are designed to run at far higher temps than most of us run them. Odds are your components will be obsolete before they fail do to heat. I have had my 2.8c up to 3.45 and it still did not get up 60c with stock cooling after a couple hours of prime 95. When I tried 3.5 It became unstable but temps were still not high. I think I just hit a wall .

 

[boyRacer]

It shouldn't but my brother's rig... after he's been playing for about 15-30 mins or so... i swear it loses 10-15fps... it's always like this though... weird.

 

[s0ssos]

actually, i think cpu's do perform better at lower temperatures.
at the transistor gate level, you see that resistance = greater load = slower pull-up time. so if it's hot, that must means there's a lot of resistance. cooler means less resistance

 

[Twista]

Originally posted by: boyRacer
It shouldn't but my brother's rig... after he's been playing for about 15-30 mins or so... i swear it loses 10-15fps... it's always like this though... weird.

how much ram does he have and what game is it? Maybe hes accessing the pagefile too much

 

[boyRacer]

Originally posted by: Twista

Originally posted by: boyRacer
It shouldn't but my brother's rig... after he's been playing for about 15-30 mins or so... i swear it loses 10-15fps... it's always like this though... weird.

how much ram does he have and what game is it? Maybe hes accessing the pagefile too much

1gb... and he was playing planetside at the time. i don't think virtual memory has any effect on fps.

 

[boyRacer]

Originally posted by: s0ssos
actually, i think cpu's do perform better at lower temperatures.
at the transistor gate level, you see that resistance = greater load = slower pull-up time. so if it's hot, that must means there's a lot of resistance. cooler means less resistance

you're right...

here's what asetek has to say... makers of vapochill:

Using a vapor compression cycle to attain evaporator temperatures in the range ?30ºC to 0ºC the CMOS core temperature in VapoChill®-based PC systems will stabilize around -15ºC to 10ºC during continuous maximum load (dependent of Vcore). The primary technical advantages to CMOS technology under sub zero temperatures are:

1. Faster switching times of semiconductor devices.
2. Increased circuit speed due to lower electrical resistance of interconnecting materials.
3. Reduction in thermally induced failures of devices and components.

Link

 

[0roo0roo]

heats just wasted energy

 

[fredtam]

Originally posted by: boyRacer

Originally posted by: s0ssos
actually, i think cpu's do perform better at lower temperatures.
at the transistor gate level, you see that resistance = greater load = slower pull-up time. so if it's hot, that must means there's a lot of resistance. cooler means less resistance

you're right...

here's what asetek has to say... makers of vapochill:

Using a vapor compression cycle to attain evaporator temperatures in the range ?30ºC to 0ºC the CMOS core temperature in VapoChill®-based PC systems will stabilize around -15ºC to 10ºC during continuous maximum load (dependent of Vcore). The primary technical advantages to CMOS technology under sub zero temperatures are:

1. Faster switching times of semiconductor devices.
2. Increased circuit speed due to lower electrical resistance of interconnecting materials.
3. Reduction in thermally induced failures of devices and components.

Link

If you strap a vapochill on a 2.8C it is still going to run 2800Mhz until you overclock it. Once you overclock it vapochill is only going to be effective if your CPU is becoming unstable due to heat. Read between the marketing. They can lhelp you achieve between 20 & 60% speed gain. I would like to know which ones besides the engineering samples were able to achieve 60% Most people on this form hit a wall at around 125% of there rated clock due to the chip itself not heat. If the heat is the problem then yes vapochill would be the answer. Besides the question referenced water cooling and not phase change systems.

 

[boyRacer]

Originally posted by: fredtam

Originally posted by: boyRacer

Originally posted by: s0ssos
actually, i think cpu's do perform better at lower temperatures.
at the transistor gate level, you see that resistance = greater load = slower pull-up time. so if it's hot, that must means there's a lot of resistance. cooler means less resistance

you're right...

here's what asetek has to say... makers of vapochill:

Using a vapor compression cycle to attain evaporator temperatures in the range ?30ºC to 0ºC the CMOS core temperature in VapoChill®-based PC systems will stabilize around -15ºC to 10ºC during continuous maximum load (dependent of Vcore). The primary technical advantages to CMOS technology under sub zero temperatures are:

1. Faster switching times of semiconductor devices.
2. Increased circuit speed due to lower electrical resistance of interconnecting materials.
3. Reduction in thermally induced failures of devices and components.

Link

If you strap a vapochill on a 2.8C it is still going to run 2800Mhz until you overclock it. Once you overclock it vapochill is only going to be effective if your CPU is becoming unstable due to heat. Read between the marketing. They can lhelp you achieve between 20 & 60% speed gain. I would like to know which ones besides the engineering samples were able to achieve 60% Most people on this form hit a wall at around 125% of there rated clock due to the chip itself not heat. If the heat is the problem then yes vapochill would be the answer. Besides the question referenced water cooling and not phase change systems.

I just posted an example of whether low temperatures have any effect on a CPU at the same clockspeed but at different temperatures. I have no knowledge regarding this issue but I am very curious. Of course vapochill affords further overclocking headroom but I want to know if you could actually improve a CPUs performance just by temperature.

Here is another article which has the same excerpts from the Asetek site.

 

[KillaBong]

A 1700+ can get at least a 60% overclock.

 

[Jeff7]

Originally posted by: boyRacer
It shouldn't but my brother's rig... after he's been playing for about 15-30 mins or so... i swear it loses 10-15fps... it's always like this though... weird.

Could be a memory leak too.