Intel Haswell CPU & Z87 chipset

[QUOTE=JReynolds;2689294]Good choice on the mobo - so neat and tidy looking :cool:.

Of course Asus can’t spell - should be sabREtooth.

Have fun m8 :slight_smile:

Regds, JR[/QUOTE]

Welcome to English. We hope you have fun trying to figure it out. -All your fellow English-speakers of the world.

tada
:smiley:


Your vendor is fast with shipping. ( again with the air cooler) I hope these don’t run as hot as the 3770Ks. LOL.

[QUOTE=alan1476;2689371]Your vendor is fast with shipping. ( again with the air cooler) I hope these don’t run as hot as the 3770Ks. LOL.[/QUOTE]

Reports from some guys is that they do, since they have the same paste under the heat spreader that the 3770K’s have.

As Jimbo says, the reports say they do run as hot as Ivy Bridge.
A lot of people don’t seem to understand that regardless if the cooling is air, or closed loop water, there is little difference between the two. In fact, IMO a closed loop water cooler on Ivy Bridge or Haswell is just wasting money.

The problem with Intel’s Ivy and Haswell is.
Transferring the heat from the CPU die to the integrated heat spreader on the CPU. Once the heat is transferred to the heat spreader, then the heat is easy to get rid of. And a good quality air cooler is just as good as closed loop water.

To get heat under control with a heavy overclock, you have to delid the CPU and replace the TIM, or remove the integrated heat spreader completely.

My Ivy Bridge 3770K runs at 4.6GHz on air quite easily.

BTW: With closed loop water cooling, or air cooling. Temps can never be lower than room temp, as you’re simply cooling the water in the radiator with air from the room. the very same air that blows through the fins on your heatsink with an air cooler.

You are correct again Wendy, the ambient temps here where I live are about 32.2 c or 90F.

To convert Fahrenheit into Celsisus,

(F - 32) x 5/9 is the formula.

So, 90 - 32 = 58 and 58 x 5/9 = 290/9 = [B]32.2 C[/B] My 3770k IS AT 39C RIGHT NOW.:sad:

[QUOTE=Dee;2689396]As Jimbo says, the reports say they do run as hot as Ivy Bridge.
A lot of people don’t seem to understand that regardless if the cooling is air, or closed loop water, there is little difference between the two. In fact, IMO a closed loop water cooler on Ivy Bridge or Haswell is just wasting money.
[/QUOTE]

It looks like if you don’t delid a 3770K, it looks like there isn’t much difference:

However, I’ve seen reviews in which closed loop WC cools better than air cooler, especially when the chip is overclocked.

Example 1 100% load, top graph:

Example 2 100% load(4Ghz) and OC’d graphs(4.4Ghz)

It all depends on how well the cooling mechanism can deal with the heat thrown at it.

The problem you face either way is…well, you have to eventually get rid of the heat from whatever is in contact with the chip itself.

A given volume of water can [subjectively] hold more heat energy than air of the same volume, so with enough room temperature water rushing directly to the hot chip, you’ll get a slightly cooler chip than if it is stuck next to a consistently warmer piece of metal.

But to a point, that doesn’t even matter, which is why air cooling is still doing so well in the market. Many of the high-end air cooling mechanisms are designed well enough that the heat is shipped off to a great volume of air as fast as the heat is produced, leaving the metal/other materials functionally as cool as water that has been recirculated a few times in a water loop.

Hi, Dee, What computer case will you install the 212 EVO cooler master in? Because of the height of it, I want to order the cooler master myself, but I don’t think that it will fit my computer case.

Thanks
Mr. Bill

Removing heat also it very dependent on the interface between the heat source and the heat sink. One thing water has going for it is a very good thermocouple to the CPU housing. The liquid is in full contact with the CPU heat sink and uses the entire surface to transfer heat. With air cooling the best heat transfer occurs when there is metal to metal contact. Most applications of heat sinks to CPU cover for air cooling has little metal to metal contact thus the need for thermal paste. Thermal past fills the voids and allows heat transfer but not as efficiently as if the two metal surfaces were perfectly flat and in full contact with each other.

When I was overclocking a E6700 CPU I read about lapping the heat sink and CPU mating surfaces and decided to try it. I was amazed at how non-flat the two surfaces were. Especially the CPU cover. I spent 2-3 hours lapping the heat sink and CPU and when I was done the CPU had a mirror finish as the cover was made of copper with a nickel plating over it. The lapping removed all the nickel plating. The result were fantastic. I was able to run that CPU at 3.5 ghz at less than 60C with a $10 aluminum heat sink with a 92mm fan. It wasn’t a heat pipe setup it was just a chunk of aluminum with fins machined into it. I haven’t lapped any other setups but if someone is looking to really push things to their limit then it is the way to go with air cooling. Here is what a lapped 3770K looks like:

http://forums.anandtech.com/showthread.php?t=2261219

Also, I have found the air coolers that have the heat pipes in direct contact with the CPU cover are more effective at cooling. It makes sense because there is less mass between the heat source and the cooling medium.

[QUOTE=Mr.Bill;2689413]Hi, Dee, What computer case will you install the 212 EVO cooler master in? Because of the height of it, I want to order the cooler master myself, but I don’t think that it will fit my computer case.

Thanks
Mr. Bill[/QUOTE]

If you look at the specs on Newegg, the Evo is 6.26inches/159mm. So you are probably going to need a case with a depth/width of at least about 6.75 inches(allowing for motherboard to case depth and some room between the case and the top of the heatsink.

Because my Dell case was small enough, I couldn’t fit the Hyper 212 in it. So I opted for a Zalman 9500AT2, which is all copper. It’s smaller (125mm/4.92 inches)and yet it does a very good job of heat transfer. One of the only downsides is that there is no way to hang a separate fan like the Hyper 212/212 EVO.

[QUOTE=yojimbo197;2689419]If you look at the specs on Newegg, the Evo is 6.26inches/159mm. So you are probably going to need a case with a depth/width of at least about 6.75 inches(allowing for motherboard to case depth and some room between the case and the top of the heatsink.

Because my Dell case was small enough, I couldn’t fit the Hyper 212 in it. So I opted for a Zalman 9500AT2, which is all copper. It’s smaller (125mm/4.92 inches)and yet it does a very good job of heat transfer. One of the only downsides is that there is no way to hang a separate fan like the Hyper 212/212 EVO.[/QUOTE]

Thanks Yojimbo for the information. But I want to use the EVO.

Mr. Bill

The case I will be using is an Antec Performance One P280, but the 212 EVO also fits an Antec 100.
The really neat thing about the 212 EVO is it uses a direct contact heatpipe design, so the heatpipes themselves are in direct contact with the CPU heat spreader.

The problem with Ivy and Haswell is they use a thermal paste as in interface between the CPU die and the heat spreader, whereas in Sandy Bridge and previous generations of CPU’s they soldered the CPU die to the heat spreader. Solder is much more effective at transferring heat compared to a thermal paste.

Also on Ivy and Haswell, you have a much smaller CPU die, and therefore a smaller surface area to make contact with the heat spreader.

So for Ivy and Haswell, you simply can’t get the heat from the CPU die to the heat spreader fast enough. Once the heat is at the heat spreader, it’s very easy to get rid of the heat.

Perhaps air cooling is so effective here because Scotland doesn’t generally have high ambient temperatures.

Thank you Dee.

Mr.Bill

[QUOTE=Dee;2689429]
Perhaps air cooling is so effective here because Scotland doesn’t generally have high ambient temperatures.[/QUOTE]

:iagree::slight_smile:

I wonder what working liquid they have in the heat pipes. Heat Pipes are a clever bit of design - well worth a google.

Regds, JR

[QUOTE=JReynolds;2689484]:iagree::slight_smile:

I wonder what [B]working liquid they have in the heat pipes[/B]. Heat Pipes are a clever bit of design - well worth a google.

Regds, JR[/QUOTE]

Distilled Water LOL;)

[QUOTE=alan1476;2689500]Distilled Water LOL;)[/QUOTE]

My money is on whisky (it’s readily available from your local distiller in Scotland)

Regds, JR

I guess what’s in! or what’s coming now its the mini-itx form factor with all the goodies of ( or almost all ) all the other form factors :cool:

Water cooling is not a heat sink solution, it’s a heat transport solution. With a large reservoir, it will buy your cpu more operating time at a bigger tdp, but it’s not necessarily sustainable.

It’s the same with heat pipes. They are not a cooling solution, they are a heat transport solution.

Both water cooling and heat pipes require an effective heat dissipation solution to dissipate the heat into the ambient air.

If the actual heat sink fins are unable to dissipate the heat quickly enough, the water in the heat pipes evaporates, and the copper pipe has to transport all the heat to the fins for dissipation. The copper pipe has orders of magnitude more heat impedance than a heat pipe working at its rated tdp.

@JR, alcohol evaporates at very low temperatures, and is therefore useless as a heat pipe coolant.
Evaporated alcohol is a great way to get drunk fast though :wink: Through your lungs and straight into your bloodstream … No waiting for the slow as molasses stomach to do it’s thing…

Thanks - I understand Debro.

Am I right to think that if a cooling solution can dissipate all of the heat generated by the CPU the rate at which heat can be transported then becomes the key factor - and in turn this is where water cooling does have an advantage over a relatively passive heat pipe arrangement.

Regds, JR