Well, I don't care about restriction in my loop since I only have 1GPU in my loop, nothing else....heh heh heh.
I do care about the corrosion factor though....and this is prime picking for corrosion.
Printable View
Well, I don't care about restriction in my loop since I only have 1GPU in my loop, nothing else....heh heh heh.
I do care about the corrosion factor though....and this is prime picking for corrosion.
Yeah but look at all those damn chips that need to be cooled. I wonder if the Mosfets would work ok with just ramsinks...Quote:
Originally Posted by wangerin
They do work fine with stock air coolerQuote:
Originally Posted by J-Mag
Yeah but when replacing stock HSF with a block, airflow will be compromised. Obviously this can be taken care of with additional fans, but that kinda defeats the purpose doesn't it?Quote:
Originally Posted by Cooper
Looks awesome but we all know that looks means **** all.
Alu
high restriction
will probably be impossible to get hold of
not good :(
gonna stick with stock coolers anyway, they look cool :D
It`s certainly does, but I can`t imagine water cooled rig w/o airflow at all. It`s simply will fry :DQuote:
Originally Posted by J-Mag
G80 is pretty powerhungry card and power circuit must have decent cooling - good ramsinks with silent fans should do the trick just fine...but then again there isn`t any GPU water block for this card atm :)
Hmm, maybe that could be because the last 5-6 waterblocks posted in this area ARE ass?Quote:
Originally Posted by Kunaak
This one's aluminum, and super restrictive, the Maze 5 is useless since it "might" perform 1-2degrees better than the Maze4/MP-01/MCW60, and what else was there? Some :banana::banana::banana::banana:ty koolance blocks? It's not the instant reply to new waterblocks, it just so happens all the new waterblocks have sucked, don't shoot the messenger.
Yeah, watercooling people loved the MCW60...:D
the block looks huge!
do I even need to point out what a dumbass design that is..
I want to see what you have to say...
No, you'll get in trouble for being negative and/or critical! :p:Quote:
Originally Posted by MaxxxRacer
looks hot, great mill work, but poor design imo.
Is it me or are the channels in the copper section running the wrong direction?
Jeez, I thought I was the only one who thought this! :2cents:Quote:
Originally Posted by Kunaak
Let's all chant now:
"It's restrictive, It's restrictive, It's restrictive, It's restrictive, It's restrictive, It's restrictive, yada, yada, yada" :slapass:
Quote:
Originally Posted by MaxxxRacer
Yes please :)
Where exactly does the water flow into the copper fins? It looks like the water just stops.:confused: Also copper + aluminium? Why? Why?
1. aluminum + copper = galvanic corrosion
2. aluminum and copper seal. the galvanic corrosion will happen fastest here, causing a leak right on the core of the gpu. suffice to say they will have some very unhappy customers if they dont supply some hardcore corrosion inhibitors with the block.
3. the gpu core cooling area. whatever moron designed this should be shot in the face. first off lets point out the fact that the fins are turned in the WRONG direction in the picture. now lets assume they will be turned the right way when you use the block. if you look closely at the edge of the copper/aluminum seam you will notice that the aluminum is cut down slightly to the height of the copper fins. this is so that the water can flow past the fins.. and that is exactly what it will do. the slit is at the top of the fins which means the base of the fins will get little in the way of turbulence (which is required for good performance) and thus there will be stagnat water there. this stagnant (relative to the rest of the water) water will heat up causing poor cooling performance. If my explanation didnt make sense lmk and i will do a paint picture
4. the channels in the aluminum. You have got to be kidding me right... nearly every full cover waterblock does this (barring a few). they make these little tiny channels that look like seismograph reading from Los Angeles in order to cool the ram effectively. Get a brain you moron block designers. The ram isnt putting out that much heat and a larger channel with lower velocity water that swept past the general vecinity of each ram chip would be MORE than enough.
5. the complex design of the gasket groove. The design of the gasket groove does nothing but increase the cost of the waterblock. If they brough the groove inside of the mounting screws and didnt have it hug the edges the block would be easier to machine, a shorter gasket could be used and the gasket would be a hell of alot easier to reseat after cleaning the block.
In short, this block was designed by someone who has little understanding of economics or thermodynamics. But hey.. they really know their solidworks :rolleyes:
It's big and it's chunky just like me .. whether it is as useless as me in general though remains to be seen .... benefit of the doubt until someone actually tests it ?
Regards
Andy
Hmnn.. is the Maze4's base big enough to cover the IHS of the G80? If it is isn't... would it still be able to cool it enough. I recall watercooling the 5900 with an MCW50... and the base of the MCW50 was not nearly the size of the 5900 and it worked fine.
The G80 is a different beast though.
Don't wanna spend 200 dollars on waterblock... especially one such as this.
The maze 4 would work OK I expect, the problem is finding a way to mount it on the card. Then you still have the problem of all the other bits that need cooling. The cost in RAMsinks for the RAM, mystery chip and the power regs might mean theres no price disadvantage to go with a fullcover block, and it'd certainly look neater IMHO.
Not really.
A block will cost you about 40-50 dollars. The upgrade for a new retention bracket is probably like 10 dollars if you already have the block. And the mosfets look like they can be cooled with longer ramsinks... my only concern is whether the GPU will be sufficiently cooled (at least BETTER than stock) with a Maze 4 or MCW60.
Here in the UK 1 of Eddys blocks costs around £55-60, an MCW-60 is £30-35, and then say you need about 24/25 RAM sinks to cover the RAM, power regs and the mystery chip. 8 Swiftech MCW-14s here are £15, so say £10 for some generic ones. Thats a total of about £60-70, so it works out the same. Although admittedly you then hit the pitfall of all fullcover blocks, you cant transfer them to other cards.
That's true. I never did like those Swiftech Ramsinks as they are expensive and they can fall off easily if you are not careful of cleaning the ram properly.
Here is how much they are.
http://www.svc.com/mc14.html
Overly expensive there are lighter and cheaper alternatives that can cool well with a low amount of airflow.
Ramsinks like these can cover multiple mosfets so you don't actually have to tack on an insane amount of ramsinks.
http://www.svc.com/crc-u01.html
And Ramsinks like these can take care of the memory:
http://www.svc.com/sac-p01.html
Then this to cover the mystery chip.
http://www.sidewindercomputers.com/miso.html
All together... it will cost a little more than half of a full size heatsink.
Oh yeh, I'd forgotten about the Southsink actually. I wonder if the CM RAMsinks you linked to are up to the job of power reg cooling, if they are then you have a viable alternative to a fullcover block, at least with US prices!
I'm not sure what you based this assumption on, I can't see it being laminar at all. With the sort of flow rates that water cooling uses and the geometry, it's definetly going to be turbulent. In fact for a 1/2" tube such as that used for water cooling, and for a flow rate of 500l/h (typical D5 lowest flowrate), that gives a reynolds number of ~18000. The transition range for laminar to turbulent occurs from 500-2000 and after that it's considered fully turbulent, and that's just for a simple straight smooth walled tube, so there's no way in the world that the flow through this block is going to be laminer. But in saying that, the mill work is very nice :DQuote:
Originally Posted by Cooper
And when considering heat exchangers, restriction isn't always a bad thing, it can help heat transfer considerably, although there it's obviously a point where it's negatives out weigh it's benefits. I seem to recall the Swiftech G5 looking and performing quite "restrictively", never heard anyone winge about it's performance though