Innovatek Watercooling Setup

1. (cpu block) Innovatek G-Flow Waterblock - socket 775
http://www.highspeedpc.com/Merchant2/merchant.mv?Screen=CTGY&Category_Code=Gflow

2. (chipset block) Graph-O-Matic Rev3.0 VGA/Chipset (Straight/Blue)
http://www.highspeedpc.com/Merchant2/merchant.mv?Screen=CTGY&Store_Code=HSPC&Category_Code=VGA3

3. (reservoir) Tank-o-Matic External (Black)
http://www.highspeedpc.com/Merchant2/merchant.mv?Screen=CTGY&Store_Code=HSPC&Category_Code=TankEx

4. x2 (video card full cover waterblock for ATI X1950XTX) Cool-Matic ATi x1800/1900 Cooler
http://www.highspeedpc.com/Merchant2/merchant.mv?Screen=CTGY&Category_Code=CoolMaticx1800

5. (pump) Innovatek HPPS PLUS 12v Pump
http://www.highspeedpc.com/Merchant2/merchant.mv?Screen=PROD&Product_Code=HPPSDC&Category_Code=Pump

6. (radiator) Need help here, I need the best triple fan radiator available.

Please give me any ideas or thoughts about this setup. I am looking for the best water-cooling setup, money is not an issue. As well is the pump a good one? Thanks for your help.

From what is commercially available to date...

CPU: StormG4, AquaXtreme MP05SP LE, or Nexxxos XP
GPU: 2x Swiftech MCW60 with MC14 ramsinks
Chipset: Don't bother. Use a Zalman passive chipset sink.
Pump: Alphacool DDC Ultra (or DDC+ with Petra modded top)
Radiator: ThermoChill PA120.3 partnered with fans to suit your noise-level preference.
Reservoir: Inconsequential. Use t-piece and fillport.

...will wipe the floor with the Innovatek setup above... read all the major forums. Looks like you've been doing more windowshopping than reading I'm afraid...

Any other ideas or opinions?

What Marci suggested is going to not only give you better performance but cost a heck of a lot less.

absolutely no question! Buy what Marci said...

If you had read a little bit you would not have even looked at Innovatek. Go with what Marci has suggested. The only additional thing is if you decide to go with a res over a t-line, get a swiftech microres.

I have read a lot on the net about Innovatek vs other companies, and most all of them are saying that they have some of the best blocks, like the g-flow. I'll take a look around again and see what others are saying.

Innovatek blocks are European styled low-flow block watercooling setups. They are geared more for looks than top performance. $125 for that g-flow is ridiculous. An $80 Storm with a good pump will blow the socks off of what that block can do, or for that matter, a $50 Apogee. Just look at their pump. It has a "whopping" 5.5' of head compared to the 20' of head of a DDC+. They also like to use anodized aluminum. If aluminum is properly anodized, it takes a long time for it to corrode if ever, but the least little gap in the anodizing will cause corrosion very quickly. It is just better to avoid aluminum at all and just go all copper so that any possible corrosion won't even be an issue regardless.

Its not that their products necessarily suck as far as quality, but if performance is what you want, go with what Marci suggested.

C/W of the block at a given flowrate is what counts. Innovatek's earlier blocks capped out at 1GPM... find out the C/W of the Innovatek blocks in question and flowrate of the pump in question, plot on the graph here: http://www.procooling.com/index.php?func=articles&disp=131

Compare the position of the plot to the selectable plots for the Storm generation blocks etc... lower the position on the vertical scale the better.

Marci, first how would you go about finding out the C/W of a block?

Second, wouldn't flow rate also matter as to the efficiency of the rad thereby effecting the cooling capacity of the entire loop? In other words if a block plots wonderfully on that chart based on its C/W rating and flow, but the system as a whole is 5c warmer because of less heat being transferred from the rad to the air due to low flow than it would if it had better flow, isn't flow still the issue? Or is 5c way out of line?

I have already had bought some Innovatek parts before making this tread. You guys were right the pump I got is small and the blades are tiny.

I got in the:

(pump) Innovatek HPPS PLUS 12v Pump
(reservoir) Tank-o-Matic External (Black)
(chipset block) Graph-O-Matic Rev3.0 VGA/Chipset (Straight/Blue)

I'm still going to order the G-flow to test it out, since there are not that many reviews on it. I also have the Stom g4 on its way as well, and a Thermo Chill Rad.

Marci, first how would you go about finding out the C/W of a block?

Click to expand...

Google lots and hope someone's done some accurate testing... or do a LOT of reading up and splash out a lot of cash to build yer own testrig. Those who formerly did such testing no longer do so due to the level of criticism from genPop.

Second, wouldn't flow rate also matter as to the efficiency of the rad thereby effecting the cooling capacity of the entire loop?

Click to expand...

Yup... but all I gave instructions for was comparing the blocks in their own rights... explaining how to work entire system C/W when the necessary figures and curves don't exist to allow one to do so is a minefield. Then we end up in the realms of testing, which brings on the whole diesim vs CPU argument etc etc.

We "know" X block beats Y block if both are used at the same specified flowrate. Just exemplifying how he can easily determine if one block is better than another using available data...

Note: Block C/W isn't calculated with a rad in the loop... generally just consists of pump and tube and block, with coolant temp controlled by recirculating chiller setup... therefore rad properties don't affect the tested block's C/W as calc'd on that testbed.

[O-CuK]Marci said:

Those who formerly did such testing no longer do so due to the level of criticism from genPop.

Click to expand...

And the sad part about that is all you are left with are potentially biased evaluations from manufacturers. For instance, how much do I trust Swiftech's testing saying that the Apogee is as good as the Storm when there is no third impartial party to either confirm or challenge this?

In your case, you have had others do testing ie radical_53 to confim your results on Thermochill rads, hence it lends a lot of credibility to your results.

I remember the big debate when systemcooling did the article in which they die tested the Apogee with other blocks and then did a "real world" test in which just about any block yielded the same results. I just didn't know that others stopped testing, although I guess I should have given that everything on procooling seems to be at least 2 years old.

Ehm guy's,
I don't want to offend anybody but i have a FULL Innovatek setup, (2x 7900GTX and an OC'ed 4400+ that is), and my ambient is 22deg. Celcius and my Proc is like 29degr. and NEVER goes over 35deg, the Vid's stay even cooler.\
And that is with a HPPS+ and a dual rad with 120mm fans...

Excalibur1982 said:

Ehm guy's,
I don't want to offend anybody but i have a FULL Innovatek setup, (2x 7900GTX and an OC'ed 4400+ that is), and my ambient is 22deg. Celcius and my Proc is like 29degr. and NEVER goes over 35deg, the Vid's stay even cooler.\
And that is with a HPPS+ and a dual rad with 120mm fans...

Click to expand...

Sweet I think I'll stick with my Innovatek setup, and if it doesn't give good results, I'll go with the Storm setup. I previously had a Innovatek setup, and it never went over 35C, and I ruined it since I was a noob when I first got it, and used tap water lol.