I7 Waterblocks Version3

[Hondacity]

Not a universal thing...definitely depends on the block and the CPU (well, really how the block relates to the CPU and how the block relates to flow).

Decided to expand my testing to the three major Intel CPU categories: dual die Core 2, single die Core 2, Core i7. Currently working on the Fuzion V2 on a blazing B1 Conroe and three mounts in on a no-nozzle setup....and I'm seeing roughly .5C difference between 3 DDC3.2s (~2.65GPM) and a single DDC3.2 at 7.65V (~.8GPM). 3.5mm nozzle should be another story of course though

where did you get .5c? i see 2.44c difference in one table..



.5c is that water temp?

[Vapor]

That's on a Kentsfield

I haven't shown any of my Conroe testing yet

(also redid the loop, removing the VL3s and remodding the MCP350s to 355s (I think one of the pumps 'lost' the mod), so flowrates should be higher now)

[Hondacity]

That's on a Kentsfield

I haven't shown any of my Conroe testing yet

(also redid the loop, removing the VL3s and remodding the MCP350s to 355s (I think one of the pumps 'lost' the mod), so flowrates should be higher now)

and so higher flow benefits kentsfields which are quadcores...

conroes are dual cores..and they don't make much heat...

thanks for testing

[Vapor]

Preliminarily, I think it's more of how the Fuzion base behaves with dual die vs. single die. Conroe is at higher settings than my Kents and is also a very high leakage Conroe, so heat output is surprisingly close.

[Hondacity]

Preliminarily, I think it's more of how the Fuzion base behaves with dual die vs. single die. Conroe is at higher settings than my Kents and is also a very high leakage Conroe, so heat output is surprisingly close.

how did you evaluate it to have high leakage? power from the wall? or temperatures?

[Vapor]

My E6700 gives my single stage a few problems when a known-normal B2 X6800 didn't at higher voltage and MHz

Subzero cooling is pretty telling for that kind of thing fortunately.

[Hondacity]

thanks

i'll wait for your i7 testing

[Vapor]

thanks

i'll wait for your i7 testing

i7 will be awhile, still finalizing the testbed....not sure what kind of CPU I want to use (C0 vs. D0, regular vs. high-leakage, ES vs. retail, etc.).

Then I'll do the Fuzion V2 gamut to make sure all is 'well' with the testbed....and then I'll work through various other CPU blocks progressively (i7 will get a higher priority than it does now though).

[Vapor]

i7 920, 1.3v, Single PA 120.3, ZM-F3's @ 1250 RPM. As far as what others are showing, can't tell ya. Proofs in the pudding and the numbers represent their selfs. CPU only loop, not other restriction in the loop except the block.

andyc

I think your results are going to be pretty reasonable and logical....you have a block that's showing signs of being flow-agnostic (too early to tell, but the signs are there) and a loop that has much less cooling power than my setup* and pumps that are more thermally expensive**.

* PA120.3 + 3x 1250RPM F3s vs. MCR320+220 sandwich with 2600RPM shrouded San Aces

** D5s dump more heat than DDCs (especially because, unfortunately, it seems my res-top DDC lost its 3.1->3.2 mod at some point during testing (probably very early, like mid-test on the Stock nozzle based on apparent flowrates) and therefore was running even cooler).

[Vapor]

Damn Andy, you do have quite a bit to test

I'll be very interested in D5s vs. DDC3.25s

[Hondacity]

I think your results are going to be pretty reasonable and logical....you have a block that's showing signs of being flow-agnostic (too early to tell, but the signs are there) and a loop that has much less cooling power than my setup* and pumps that are more thermally expensive**.

* PA120.3 + 3x 1250RPM F3s vs. MCR320+220 sandwich with 2600RPM shrouded San Aces

** D5s dump more heat than DDCs (especially because, unfortunately, it seems my res-top DDC lost its 3.1->3.2 mod at some point during testing (probably very early, like mid-test on the Stock nozzle based on apparent flowrates) and therefore was running even cooler).

it also explains his water temp difference between 1pump and 2pumps ..if it were a better rad the temps would have been better..and not almost the same...

[Hondacity]

Tell ya though,

I'm about to get the flow rate monitoring itch So I might just have to do something about that one

KL's back on the bench now, just filled the loop. Man what a difference in flow rate. I have to get something so I can see exactly just how much difference.



Good theory, but not conclusive by any means. Think about it, everything's relative to one another in a loop. If I would have had more rad, the temps would have been better to begin with. Regardless of how much cooling capacity you have, the ability to remove X amount of heat is fixed all things being equal. If you induce more heat dump in the loop, temps go up all things being equal. So the theory that more rad or cooling capacity somehow offests handleing the 2nd pumps heat dump in the loop, doesn't hold up.

It's like saying that turning the fans up higher to negate the greater Delta T will offset the difference and you'll see a gain with two pumps. I could have just saved myself the trouble along with the extra pump, and tuned the fans up higher to begin with. Same results.

andyc

interesting

i await your results

[BringerOdeath]

Hmmm, I have 2 pumps in a single loops with three rads and I get better temps on everything than when I had 2 loopsand this is with only 2 fans on the 480, 1 fan on the 240, 3fans on the 360...I am using one d5 and a D4, I am cooling the cpu, mosfets, NB/SB, and dual GTX 295's.... I decided to get the HK 3.0 LT for the lower Pressure drop and the temps are very close! Could the unpolished base be the temp difference between the KL 350 and HK?

[Hondacity]

hahahahha if it were polished i would have given that build quality score a 10...but yeah the thermal performance is really close to that of the k350...i just wish it were polished...

[Bladesinger7x]

Great test Hondacity ...how is EK Supreme comparing to the HK block?

I need to make a decision about the water block in the following days,
and so far Supreme or HK Rev.3.0 LC are the only option to get without
paying an arm and a leg for the damn block.

[Hondacity]

I can't say yet...i ran out of mx2

[Sadasius]

I can't say yet...i ran out of mx2

Say it ain't so....If your doing testing then this is what you need...

[Hondacity]

i got some coming...all i did this weekend was shopping