AMD/ATi HD4850/4870 Fan Fix!

[Bo_Fox]

Yeah, ATI should have done that in the first place. 30% speed rather than 10% at idle would still be very quiet and keep it cooler running, and like 65% at load (with a 55-60 degree target limit) would actually REDUCE the power consumption of the chip (besides increasing the longevity and overclockability).

The power consumption is not that good with 4870's, actually exceeding the 8800GTX by a few watts under max load. If it was kept at 55 or 60 degrees, that would shave at least a dozen watts off the chip, as there would be less power leakage.

On Penryns, a 20 degree difference saves like more than 40 watts when feeding that thing 1.5v, and does wonders for overclockability.

[josty2]

Yeah, ATI should have done that in the first place. 30% speed rather than 10% at idle would still be very quiet and keep it cooler running, and like 65% at load (with a 55-60 degree target limit) would actually REDUCE the power consumption of the chip (besides increasing the longevity and overclockability).

The power consumption is not that good with 4870's, actually exceeding the 8800GTX by a few watts under max load. If it was kept at 55 or 60 degrees, that would shave at least a dozen watts off the chip, as there would be less power leakage.

On Penryns, a 20 degree difference saves like more than 40 watts when feeding that thing 1.5v, and does wonders for overclockability.

That's just nonsense. The card does NOT consume less power if it is cooler...
It will consume maybe 1W more, because the fan spins faster, but it will never start producing less heat by being cooler, it can just get rid of the heat easier, making it cooler...

[Shadowmage]

That's just nonsense. The card does NOT consume less power if it is cooler...
It will consume maybe 1W more, because the fan spins faster, but it will never start producing less heat by being cooler, it can just get rid of the heat easier, making it cooler...

This is actually incorrect... it's well known that transistors leak more current (static leakage) with increases in temperature.

[Bo_Fox]

That's just nonsense. The card does NOT consume less power if it is cooler...
It will consume maybe 1W more, because the fan spins faster, but it will never start producing less heat by being cooler, it can just get rid of the heat easier, making it cooler...

As your mentor, I tell you, "Go and do some research."

[eleeter]

If we follow Ohm's law, the power consumption of the card will not change depending on temperature, assuming a consistent load.

Or to rephrase it slightly, the resistance of a graphics card stays nearly the same regardless of heat, which must mean the same power draw.

The way I understand it is, when powered on initially, the card absorbs the energy first. When it reaches it's static output/cooling point, the heat is then transferred to the surroundings. If you increase the fan speed, it means the heat is more quickly transferred to the surroundings on start up, and of course this is maintained, and the cards terminal temperature is lower. But the total heat output does not change, in fact it cannot change because it must obey Ohm's law.

But then again, I have quite a bit of electronic experience, but I am no thermal expert so I could be wrong. But to me, because the electrical energy is being turning nearly 100% into heat, it makes sense.

[Bo_Fox]

If we follow Ohm's law, the power consumption of the card will not change depending on temperature, assuming a consistent load.

Or to rephrase it slightly, the resistance of a graphics card stays nearly the same regardless of heat, which must mean the same power draw.

The way I understand it is, when powered on initially, the card absorbs the energy first. When it reaches it's static output/cooling point, the heat is then transferred to the surroundings. If you increase the fan speed, it means the heat is more quickly transferred to the surroundings on start up, and of course this is maintained, and the cards terminal temperature is lower. But the total heat output does not change, in fact it cannot change because it must obey Ohm's law.

But then again, I have quite a bit of electronic experience, but I am no thermal expert so I could be wrong. But to me, because the electrical energy is being turning nearly 100% into heat, it makes sense.

I cant remember where there was such a wonderful article showing the numbers and real-world benchmarks, but here's a direct quote from an Intel guy saying:

Matty @ Intel:

Yes, the power consumption is reduced when the temperature of the processor is lowered.

There are many things that happen in a CPU when the temperature is changed and to elaborate further on the processor specific causes we have to look at the origin of the power consumption. We can divide the total consumed power into two main parts, static power (Ps) and dynamic power (Pd).

The static power consumption is what we usually call the leakage. In an ideal transistor, it should completely shut off the channel between the source-drain, gate-source and gate-drain. Transistors are far from ideal, and the current leaks between these parts and the substrate of the processor, and this is heavily dependent on the temperature.
For example, going from room temperature to 85C (~60C difference) increases the leakage power by a factor of more than 50. Thus, reducing the temperature with the same amount will make a huge impact on Ps.

Dynamic power consumption is emitted during the short amount of time that the transistor switches. Lower temperature reduces the resistance in the processor which results in shorter delay/faster switching of the transistors. Shorter delays and less noisy signals also reduce Pd.

I hope this explanation give you some clarity to the relation between power consumption and temperature. This can even be seen with air cooling: The power consumption is lower just after a load is applied compared to after a while when the temperature has levelled out, even though the load is the same.

Here, a link from googling it:
http://www.analog-europe.com/howto/208700231

[Neuuubeh]

I cant remember where there was such a wonderful article showing the numbers and real-world benchmarks, but here's a direct quote from an Intel guy saying:



Here, a link from googling it:
http://www.analog-europe.com/howto/208700231

interesting tidbit :p. So keeping a cpu/gpu/whatever as cool as possible actually reduces the heat output? so its actually kinda cooling itself so to speak

sorry for offtopic..

[sxr71]

If we follow Ohm's law, the power consumption of the card will not change depending on temperature, assuming a consistent load.

Or to rephrase it slightly, the resistance of a graphics card stays nearly the same regardless of heat, which must mean the same power draw.

The way I understand it is, when powered on initially, the card absorbs the energy first. When it reaches it's static output/cooling point, the heat is then transferred to the surroundings. If you increase the fan speed, it means the heat is more quickly transferred to the surroundings on start up, and of course this is maintained, and the cards terminal temperature is lower. But the total heat output does not change, in fact it cannot change because it must obey Ohm's law.

But then again, I have quite a bit of electronic experience, but I am no thermal expert so I could be wrong. But to me, because the electrical energy is being turning nearly 100% into heat, it makes sense.

I don't know about silicon but don't metals become more resistive with temperature? I know that bipolar transistor output devices can go into what they call thermal runaway where the heat causes the resistance to increase and hence to provide the same output the device must dissipate more heat and that causes a cycle of more and more heating to eventual failure. I understand that Mosfets are immune to this problem somehow.

Now making a device that is rated to run at 80C to run at 40C probably won't make a big difference. I'd get an aftermarket cooler to increase headroom for OC and to reduce noise, not so much to brag about my temp at idle.