Radiator / Fan Change

I'm afraid to say that zoomee was absolutely correct.

Despite a pretty large upgrade to my water cooling (new block, new larger/deeper radiator, new res, new liquid) I've not been able to improve temps really.

With a deep 420 rad and 3 high CFM fans in push only and having added my ATi 6990 to the loop I'm maintaining very similar temps to before and when CPU is under 100% load the rad does not seem to get all that hot and nor does the temps on my GFX card get much higher.

To conclude this project I've just bought a Xigmatek Elysium to house my 420 rad as despite my every effort, I couldn't cut a big enough whole with my dremel to fit it in properly (cage for 5 1/4" stuff gets in the way).

I'll also be adding the 140.1 rad I bought previously (but have yet to fit) in between my GPU and CPU.
 
Tbh, those 6990 are pretty hot... I'd not be too upset.


Need someone to make a much more effective block clearly
 
axer said:
I'm afraid to say that zoomee was absolutely correct.

Despite a pretty large upgrade to my water cooling (new block, new larger/deeper radiator, new res, new liquid) I've not been able to improve temps really.

With a deep 420 rad and 3 high CFM fans in push only and having added my ATi 6990 to the loop I'm maintaining very similar temps to before and when CPU is under 100% load the rad does not seem to get all that hot and nor does the temps on my GFX card get much higher.

To conclude this project I've just bought a Xigmatek Elysium to house my 420 rad as despite my every effort, I couldn't cut a big enough whole with my dremel to fit it in properly (cage for 5 1/4" stuff gets in the way).

I'll also be adding the 140.1 rad I bought previously (but have yet to fit) in between my GPU and CPU.
Click to expand...

It's all fun anyway!

It's a common misconception that adding more rad will automatically decrease component temps, there comes a sweet spot where the sheer heat output of a component will exceed the loops capacity to remove it quickly enough.
 
I wonder where the limiting factor is...

Whether it's the blocks ability to move the heat into the water or if more water would work
 
dazzerd said:
It's a common misconception that adding more rad will automatically decrease component temps, there comes a sweet spot where the sheer heat output of a component will exceed the loops capacity to remove it quickly enough.
Click to expand...

This is the issue with the advice most have been giving, the loop has ample capacity and hence the water doesn't even get that hot (hence the GFX card temps remain the same) the issue seems to be that I'm not able to get enough temperature out of the CPU and into the loop in the first place.
 
Which is basically because in general, cpu blocks are ****.
gpu blocks are fine becaus ethey have a much larger surface area to dispense heat unlike cpu's.

i too get high temps on my cpu whilst my gpu is cooler yet has a higher TDP then my cpu.
 
Sorry to hear you didnt have much joy mate.
Its just as i suspected though. We have to be realistic and your findings are invaluable to us all mate so thanks.

Realistically even though these 2700s can handle 1.4v plus the heat generated as a result is unsustainable for long term 24/7 useage. 5ghz is just too much for the sb architecture tk handle in its current form comfortably.

Even 4.8 (1.35-ish volts) starts pushing towards comfortable heat limits.
Thanks for sharing with feedback axer i ended up with a big case of cba and dumped my entire watercooling and have now gone back to a piddly corsair h100 at 4.6 pending ivybridge for now
 
Your welcome, hopefully the extensive testing I've done will stop all the "well my 920 runs hotter than yours yada yada" nonsense as these 2700's don't seem to perform as expected :)

Bottom line, unless you get lucky, don't expect to see sub 70degC under IBT regardless of what 120.3+ loop you run and expect that to increase if you have any GFX cards in the mix!

My intentions now to see how things are post case change and the extra RAD being put in between GFX and CPU, I'll probably drop to 4.8Ghz @ 1.35 though.

One things for sure, I certainly don't miss the sound of my ATi 6990 fans when playing BF3!
 
Last edited:
How could we make better cpu blocks then?

They already have a pretty thin base don't they... Copper has a fairly good thermal conductivity approximately 600 times that of water. Presumably the limiting step for raising the temperature of the water is then the rate at which you can transfer the heat into the bulk of the water.

With a higher restriction block (say with a one piece copper heatsink in it to massively increase the surface area, or even multiple little pipes) improve the rate of the heat transfer sufficiently?

Essentially to improve things we need to look at how to heat up the water more efficiently from the cpu
 
axer said:
Your welcome, hopefully the extensive testing I've done will stop all the "well my 920 runs hotter than yours yada yada" nonsense as these 2700's don't seem to perform as expected :)

Bottom line, unless you get lucky, don't expect to see sub 70degC under IBT regardless of what 120.3+ loop you run and expect that to increase if you have any GFX cards in the mix!

My intentions now to see how things are post case change and the extra RAD being put in between GFX and CPU, I'll probably drop to 4.8Ghz @ 1.35 though.

One things for sure, I certainly don't miss the sound of my ATi 6990 fans when playing BF3!
Click to expand...

is that aimed at me :rolleyes:
 
Ok cool but many have given facts that a 120.3 rad is more then enough to cool your cpu. There was even an exact replicated thread in another forum that came to the conclusion that these cpu's just run too hot at that clock speed :(
 
jonneymendoza said:
There was even an exact replicated thread in another forum that came to the conclusion that these cpu's just run too hot at that clock speed :(
Click to expand...

That's exactly the issue, if you look through this thread and the thread dedicated to the 2700k, people who don't have one of these CPU's but see the temps are making contrary statements of *fact* which are just guestimates at best.

I appreciate everyone wants to chip in when they think they are helping though.

Anyway, hopefully my huge waste of time and money will save someone else a lot of aggro :D
 
axer said:
That's exactly the issue, if you look through this thread and the thread dedicated to the 2700k, people who don't have one of these CPU's but see the temps are making contrary statements of *fact* which are just guestimates at best.

I appreciate everyone wants to chip in when they think they are helping though.

Anyway, hopefully my huge waste of time and money will save someone else a lot of aggro :D
Click to expand...

What i am saying is that is you did not "have" to waste your time and money on something that was already confirmed by another thread/forum.

your last comment could be aimed at you too because it did not "save u" from the aggro :D:p;)
 
But there were many more threads else where suggesting that my temps were not correct and that something was amiss and hence why I decided to upgrade ;)
 
I have similar setup to the OP, EK Blocks in a HAF-X running a 2600k at 5ghz 1.46 vcore + 6990 being cooled by the loop, SR1 360(3 x apache 120mm(Push) + 2 x 200mm(Pull ) + a SR1 240(two vipers 120mm + 3cm shroud(Push)

I have the new jet plate, doesn't make a lot of difference IBT maxes at 68oC atm, need to do some more modding of the case this weekend so can draw air in the bottom of the case for the 240mm rad will let you know if the temps go down further.

Also going to try lowering pump speed on the D5 from current setting of 4, to see if can get the water to heat up up some more :)

4.8ghz at 1.35 vcore is fine but I want 5ghz at lower temps lol
 
It would be interesting to see if a different block would make a difference though. Starting to have doubts about future potential of my Supreme HF now. Did you try cutting the piece of metal out of the two slot plate and fitting that to the block? It's the same size gap as the new plate then which is supposed to increase performance.
 
sick-lizard said:
I have similar setup to the OP, EK Blocks in a HAF-X running a 2600k at 5ghz 1.46 vcore + 6990 being cooled by the loop, SR1 360(3 x apache 120mm(Push) + 2 x 200mm(Pull ) + a SR1 240(two vipers 120mm + 3cm shroud(Push)
Click to expand...

Do you have any photos of your setup?? This is very relevant to my interests :D
 
You must log in or register to reply here.
Back
Top Bottom