My WB Testing Results, Excel file inside.

[nikhsub1]

Alright guys, I think it goes without saying that testing just sucks. You get people trying to influence you from all angles, the work is tedious and boring. To tell you the truth I really lost interest as time went on - certain individuals made the whole thing 'un-fun'. I have no affiliation and don't stand to profit from any company or individual. I do want to thank those that sent water blocks, I want to thank Swiftech for their support and blocks (as well as the facility tour) and D-Tek for the blocks. Some tests you will note are either incomplete or do not have the 5 runs done. There are various reasons for this that I can outline for you. The Storm got 4 runs - it is getting ownd and I didn't see the point in continuing with it, 4 runs is all she can take. You will note the 4th storm test was done on the back end, this is due to the fact I switched from 3 tests to 5. Same goes for the MCW6002, 3 mounts for it. The important recent contenders are the Fuzion and the Apogee line as they currently lead the pack. The MP-05 LE got one mount, yeah I know but by this point I had enough - it is going to perform near identical to the Storm on an IHS 6600. You will see some nozzle testing of the fuzion as well - and a bowed fuzion with no nozzle. Interesting to say the least. And guess what? The Fuzion with the 4.5mm nozzle has LESS restriction than the Apogee GTX (thanks for that Alex!).

I'm sure you'll have questions, and my excel file is a mess.

If anyone is good at graphing in excel, please feel free to do something to make it look nice PLEASE.

I forgot to explain a bit about my testing procedures. First, the was all done with an Asus P5W-DH Deluxe and a E6600 Intel Core 2 Duo. The CPU was run at STOCK speed (2.4Ghz) but with 1.55v set in bios. Load temps were generated with TAT and temps were attained with TAT via the logging function. TAT takes a reading of each core every 2 seconds. Each cores readings are averaged, then the 2 averages are averaged to get a single average temp. FYI TAT generates MUCH more heat than any other load program period. It has been said that TAT is a 'power virus' and will actually get the chip to output about 20% more than Intel's TDC rating. Pump used was an Iwaki RD-30 at 18v, rad is a Thermochill HE (not PA!) 120.3 with 3 102 CFM sanyo denki fans in pull on a shroud. Fans for testing were always at 12v. A Swiftech micro res was used as well. Ambient air temps were taken and recorded every 10 minutes by the digital sensor in the test room. Water temps were taken by an identical sensor in the water. Only pure distilled water was used, no additives were used whatsoever. Thermal paste used for all tests was Arctic Ceramique and I use the 'dollop' method. D-tek mount hardware was used with ~50lbs of total mount pressure - plastic stops were used so every mount had the same amount of pressure. I tried to test when ambient temps were close, but this is not always possible without an environmental chamber which of course I do not have.

Here is a pic of the test setup:



*Edit for the dope graph courtesy of Marci! Seems he plotted each MOUNT, so we can easily see the mount variation of each run. VERY nice!

[chunkylover77]

Awesome work Nikhsub. I was wondering where you think a G5 fits into?

[nikhsub1]

Awesome work Nikhsub. I was wondering where you think a G5 fits into?

I will test it, I've just been way too lazy and uninspired... Cathar said he was sending me a G7 to test... Stew?????????????

[ranker]

Pretty much what was expected but thanks for validating. The really intriguing results were the GTX bowed results and the 4.5mm nozzles on the Fuzion.

It seems to imply that the Fuzion is better overall due to requiring only a slight non-permanent mod to achieve similar if not better results than the bowed GTX.

[Petra]

As a little added support... Here's some pressure drop vs. flow data for the FuZion and Apogee GTX:



The differential pressure data was gathered using my Sper Scientific 840081 manometer (and their data acquisition software) and flowrate monitored and adjusted with a Key Instruments FR4L66BVBN (+/- 3% full-scale) flow meter. I realize that this sounds like quite a miss-match (+/-3% FS vs. +/-0.3% FS, manual "data acquisition" vs. automated acquisition and logging, etc.) but it's the best I have right now. To help make up for the inadequacies of the flowmeter, each data point on the graph represents the mean average of three non-consecutive trials. The "raw" pressure data that's on 'sheet 1' of the spreadsheet file shows the rounded max/min/mean/std. dev. output of the manometer's data logging software for each trial (again, 3 non-consecutive trials for each flowrate) and the mean result of each after three trials (the sheet titled 'FuZion' contains the graph, trend lines, etc.).

[chunkylover77]

G7 is finally coming? Wish I could afford that baby.

[[XC] DragonOrta]

Looks like I'm gonna need to get another WB soon. Thanks for all the hard work you put into this, nikhsub1.

[Petra]

If anyone is good at graphing in excel, please feel free to do something to make it look nice PLEASE.

Is this what you were thinking of, Scott? (rounded mean average of your test runs)

[Scarlet Infidel]

This is excellent stuff. Ive always wondered how far behind the modern blocks my MP-05 was. Thanks.

[nikhsub1]

Is this what you were thinking of, Scott? (rounded mean average of your test runs)

Couple oddities... averaged, the FuZion 6.5mm nozzle data seems odd (borked mounts, perhaps?), wtf is the 'Apogee GT Hybrid', and I'm assuming that the 4.5mm nozzle tests were non-bowed (and that the bowed FuZion tests were with no add-in nozzle), yes?

I was hoping to have each mount in a graph, but that may be too busy? ROFL.

Actually, the 6.5 nozzle data should really be discarded... the nozzle was not inserted properly, made that mistake on the 4.5mm at first too. To clarify, the 6.5 nozzle is just the nozzle, the 4.5mm nozzle testing is with a PROPERLY inserted nozzle and the ek barb o-ring between top and mid plate - so this is nozzle + bow if you will. The 'bowed' fuzion is with no nozzle and just the oring. Now, the GT hybrid was something I had to try... I wanted to be certain that the better performance of the GTX was due to the deck height of the top and the diagonal flow arrangement and NOT due to the more square mount. I guess a picture is worth...

Oh and Alex, your Apogee GT Flat? in the graph is actually the GT bowed but mounted VERTICALLY and not horizontally... bout 1-1.5C better as you can see. Gabe mentioned that on the GT with a DUAL CORE (single die) it matters which way the water flows.

[Grinch]

nice work...*-)

[rdmty]

Gabe mentioned that on the GT with a DUAL CORE (single die) it matters which way the water flows.

what about quad core?

[Petra]

I was hoping to have each mount in a graph, but that may be too busy? ROFL.

Well, if you did a separate graph for each set of results, then it wouldn't be too bad... you'd just have a lot of graphs.

Actually, the 6.5 nozzle data should really be discarded... the nozzle was not inserted properly, made that mistake on the 4.5mm at first too. To clarify, the 6.5 nozzle is just the nozzle, the 4.5mm nozzle testing is with a PROPERLY inserted nozzle and the ek barb o-ring between top and mid plate - so this is nozzle + bow if you will. The 'bowed' fuzion is with no nozzle and just the oring. Now, the GT hybrid was something I had to try... I wanted to be certain that the better performance of the GTX was due to the deck height of the top and the diagonal flow arrangement and NOT due to the more square mount.

Chart updated accordingly....

Oh and Alex, your Apogee GT Flat? in the graph is actually the GT bowed but mounted VERTICALLY and not horizontally...

The Apogee GT (bowed, vertical) is/was already on the chart... You have a set of tests that are labeled "Apogee GTX" (between the bowed FuZion test and the last 4.5mm nozzle test) that I'm assuming are of a stock/flat GTX, no?

[nikhsub1]

The Apogee GT (bowed, vertical) is/was already on the chart... You have a set of tests that are labeled "Apogee GTX" (between the bowed FuZion test and the last 4.5mm nozzle test) that I'm assuming are of a stock/flat GTX, no?

Yes sorry you are correct... 2 mounts of the 'flat' GTX. As you can see, the performance is well... so that is why I didn't bother to continue...

@ rdmty - I don't have a quad core and I have no interest in rerunning these tests for quad ATM.

[Cathar]

Interesting. In my own testing, a single 3/16" nozzle (4.76mm), was found to be optimal for the initial Storm/G1 prototype (single jet) in terms of trading off volumetric flow for jet velocity with typical pumps. It worked out to be just a slight touch better than 4.5mm.

[nikhsub1]

Interesting. In my own testing, a single 3/16" nozzle (4.76mm), was found to be optimal for the initial Storm/G1 prototype (single jet) in terms of trading off volumetric flow for jet velocity with typical pumps. It worked out to be just a slight touch better than 4.5mm.

Very well could be... cause right around 4.5mm seems to be the spot.

[AndyM]

Thanks for the work, good to see some hard results! A few questions...

Was the CPU lapped?

FWIW, my 'flat' Fuzion actually has a slight bow to it when checking contact patterns using Coolaboratory LM under light pressure on my lapped CPU, so I am wondering if the better temperature results of the 'flat' Fuzion and bowed Fuzion along with the bowed Apogee are a result of block design or simply better contact with the center of an unlapped CPU.

Do you have any results of a 'flat' Fuzion with the 4.5 mm nozzle to compare?

[Cathar]

Very well could be... cause right around 4.5mm seems to be the spot.

The RD-30 favored slightly smaller jet sizes 'cos of its high head and moderate flow, so that's why it'd be working well. Various other pumps with less pressure need that slightly larger nozzle diameter to find the right balance, whereas the RD-30 really doesn't care so much. It's more a case of finding the "plateau" that's broad enough to cater for a variety of pumps. Doing research based on the RD-30 is prone to skew things to the more overly restrictive side of the coin, 'cos the RD-30 will just push on through restriction more easily than most pumps will.

As for the G7, I'd been housebound up until this week. Am now able to limp around with a single crutch now that the leg cast is off, and am able to drive a car again. It's a little painful (not on any painkillers which is a good thing), and I'm going to head to the machinists today to modify a Rev0 G7 prototype base-plate for you. I'll then conduct some testing of it against the Swiftech Storm, to ensure that the base-plate modifications are doing their thing properly, and assuming it all goes well, I'll send it your way. Gimme another 3 weeks or so for all that to happen. It's not the latest G7 prototype, but it'll give an idea.

[JargonGR]

If only you had published thi two days ago. Oh well now I ordered the GTX and it is heading for bowing. I am really curious anyway how much better than my original Apogee it will be.

[sick_g4m3r]

those deltas hurt my eyes!!!

basically that is what WCing will be stuck with until the IHS heat issues are resolved?

[ranker]

Very well could be... cause right around 4.5mm seems to be the spot.

Hey Nihk,

Can you provide a small little guide on how to install everything that's needed to achieve your end product of a bowed Fuzion with 4.5mm nozzles (and what/where to purchase)? I'd appreciate it.

[sick_g4m3r]

Hey Nihk,

Can you provide a small little guide on how to install everything that's needed to achieve your end product of a bowed Fuzion with 4.5mm nozzles (and what/where to purchase)? I'd appreciate it.

yeah copy that soldier. i want to know too. for the bow isnt it just an EK barb O-ring though? but what about the nozzle?

[Bun-Bun]

Awesome, this is exactly what I have been looking for. Thanks for the work nikhsub1, it is greatly appreciated

So if I am not into bowing then the Fuzion is my choice... which I already have anyway. When gabe finishes his limited edition copper top for the GTX I will still pick it up and try it out.

What are these nozzles? Are they something you have done yourself or can you buy them?

[nikhsub1]

First to answer the question about my cpu, it is UNLAPPED...

@Cathar - send that badboy over!!!! I still need to test the G5 on my test bed but that means ripping it out of my work machine... i gotta change tubing on it soon anyway so...


Here is where the ek barb oring goes - keep in mind this is a ONE WAY STREET!!! You bow any of these blocks and there is no going back! Yes the fuzion is slightly bowed (convex) from D-Tek - this makes it more pronounced and gives a bigger mount variation too - mount several times to find the sweet spot. The nozzles were prototypes made by D-Tek and sent to me - Danny keeps saying that they are in production (the nozzles) and should be ready to go. Email him and bug him, IDK why it is taking him so long.



Nozzles - 4.5mm installed with 6.5mm and 7mm on desk.

[rdmty]

@ rdmty - I don't have a quad core and I have no interest in rerunning these tests for quad ATM.

Oops was just wondering if Gabe said anything about the mounting direction for quad cores.


When you say mounting the GT vertically as opposed to horizontally, do you mean the in/outlets should be on top of each other rather than to the left/right of each other?


Thanks for the results btw! (makes me want a fuzion now)