Originally Posted by Philth
You're thinking too much about theoretic ideals and not enough about real-world results.
In large-scale cooling systems (think outside of PCs) this can be true, but think for a second the relative size of your pc loop and then the speed at which your pump is pushing that water. The water is not sitting above your block long enough to absorb enough heat to make a notable difference from when it went into the block to when it came out. The water eventually accumulates heat, but it's gradual as it reaches stability throughout the loop.
This isn't "my theory"; this is generally accepted that the water in your loop reaches an equilibrium, at which point the order in which you place blocks/rads becomes moot.
But, feel free to grab some temperature readings and see for yourself. In fact, it would be good to get hard evidence to clear up the issue for everyone.
While you've only got a CPU in the loop you are correct in saying
The water is not sitting above your block long enough to absorb enough heat to make a notable difference from when it went into the block to when it came out.
but once you've got several hundred Watts of heat being absorbed by the water, a notable increase in temperature is seen. For instance, if you load up both of my graphics cards and my CPU, the water coming out of the final block is noticeably warmer.
In a loop like mine, I'm looking for the best possible temperatures for my CPU. This is why my loop is setup in such a way. If I was to run the opposite way around with my pump and GPUs before the CPU, I'd bet that I would see a increase in both idle and load CPU temperatures.
Originally Posted by thrasherht
You realize the energy it takes to heat up water? it is a TON which is why water cooling is so effective, it can absorb a ton of heat without heating up.
Same goes for the other way around, it will also release a ton of heat without cooling down.
If you have taken Chem, it would be the specific heat, which is the amount of energy it takes to heat one gram of water, one degree centigrade. That would equal one calorie by the way.
if you look at how much heat it dumped into the loop per second by a video card, and compare that to how much water is passing through the GPU block in that same second, it would take more energy then the GPU is putting out to heat the water that much.
So, what if you happen to have a budget system? One with a low flow pump and multiple waterblocks.
water cooling is so effective, it can absorb a ton of heat without heating up.
is obviously incorrect. Think about what you are saying there. Heats go to go somewhere, right?
My three water blocks are absorbing upwards of 500 Watt of heat. For the water to be negligibly warmer, you would have to have hugely huge flow rates.
The optimum (For CPU and GPU temp) setup is this:
Res > Pump > Rad > CPU > GPU > Res
Ensures the Res is feeding the pump, the pump is pushing, not sucking and minimum heat from the pump reaches the CPU also.Edited by Willhemmens - 8/2/11 at 5:13pm