chipset blocks a dying bread?

[jkresh]

It seems that most new motherboards (high end ones at least) are moving away from active chipset cooling and towards passive (often with heatpipes running across the board to mosfets)… Any thoughts on this trend? Will it continue? Is water on a chipset going to be come more and more complicated (as the passive systems become more advanced they tend to be harder to remove) and is it even necessary? Seriously considering going to Conroe from my current setup and probably selling my dfi chipset block (along with the board), as most of the Conroe boards I have seen so far (975 and 965) are passive. However as the block is fairly nice and very low profile (no problem using it with sli/crossfire boards) if anyone thinks we are going to go back to active cooling on chipsets in the next year or two then maybe I will keep the block.

[techun]

well i dont think anything should be described a dying bread...

i think mobo makers need to move it away from the video cards, so it can have a more decent hsf.

it was never necessary to w/c it, the old fans were just noisy, cuz they had to be small

[jkresh]

True it wasnt necessary, but it did lower chipset temps about about 17C and decreased the noise quite a bit.

[Alien1099]

Cooling chipsets, hard drives, and even mosfets is silly in my opinion. Not only is it not necessary, but it just adds more heat to the loop and decreases the efficiency of your water loop, defeating its purpose. What does it matter if you dropped the temperature of your chipset? This doesn't give you any noticeable benefit in performance that I'm aware of. The factory cooler on it should work fine. Sometimes they can be loud, but you can always remedy this.

[kiljaden5]

Cooling chipsets, hard drives, and even mosfets is silly in my opinion. Not only is it not necessary, but it just adds more heat to the loop and decreases the efficiency of your water loop, defeating its purpose. What does it matter if you dropped the temperature of your chipset? This doesn't give you any noticeable benefit in performance that I'm aware of. The factory cooler on it should work fine. Sometimes they can be loud, but you can always remedy this.
Its all about the noise (or lack thereof thanks to the chipset block) for me.. and most others that watercool the chipset

[sir_pyro]

Noise is exactally why I'm installing a second loop almost exclusivly to cool the chipset. I'm upgrading a lot of stuff, and had enough parts elft over to make a loop just for that and remove the loudedt part of my computer: The NB fan

[Ankun]

I'd imagine chipset watercoolign would be much more common if A64s werent the CPU of choice and nearly all S939 motherboards have the chipset blocked by the graphics card..

If I could fit a block on my northbridge I'd put one there, under higher clocks the chipset can get up to 55C and the only solution is a fan (over my passive zalman sink for it) which = noise.

[pejcao]

No one needed to jump over the grand canyon sitting on a weird rocket, but somebody did (...or at least tryed...)

If you give a step forth, you'll see people WC'ing their ram, even companies selling RAM WC solutions, which is less critical than NBs (meant useless).

Question is: Is it needed? An "entuciasta" will swear it is, a cheap one like me will disagree; but as I discovered since the begining, WC'ing is more a religion or being in an "elite" group than a solution ;)

BTW, those heatpipes u mention, they look like copper pipes 4 water ;)

4Give my english+spell.

[Moto7451]

Cooling chipsets, hard drives, and even mosfets is silly in my opinion. Not only is it not necessary, but it just adds more heat to the loop and decreases the efficiency of your water loop, defeating its purpose. What does it matter if you dropped the temperature of your chipset? This doesn't give you any noticeable benefit in performance that I'm aware of. The factory cooler on it should work fine. Sometimes they can be loud, but you can always remedy this.

It really depends on your architecture. Intel chipsets benefit because it cools the memory controller and higher FSB speeds can be reached. On AMD, you're right, it makes no sense since the memory controller is on die and the HT link provides more bandwidth than is generally necessary for a home computer (servers are different). I usually run my HT Multi at 1x since I run a high HTT speed and I don't see any decrease in performance. All I see are lower chipset temps since I'm running at 1/3 the spec speed.

[ƒÓÒl]

I've watercooled my NB's for years and will continue to do so.
It's not so much for silence, as this DFI had a passive 'sink on it, but that little 'sink would run at near 60C...idle!.
Removing that heat via the (external) radiator is one less thing to worry about as well as one less fan involved in cooling the case temps.

Also, I'm not a rich man, and I firmly believe that a cool computer lasts longer. Period.
It doesn't hurt that I'm using a block I got for $5 from Gemini's going out of business sale. ;)

If they stop making chipset blocks that fit, I'll go back to making my own.
If they stop giving me mounting holes for them, I'll glue the darned things on!
Watercool the world!

[Gautam]

As Moto said, A64's wouldn't benefit from it in regards to noise. Most of the boards I've been using lately have had passive heatsinks, but otherwise, I would definitely do so. I think it'll become more common with Conroe. I've got an MCW30 myself that's waiting to be put to use. I bet plenty will gain overclocks with Conroe if they step up the cooling on the northbridge.

[kiljaden5]

As Moto said, A64's wouldn't benefit from it in regards to noise. Most of the boards I've been using lately have had passive heatsinks, but otherwise, I would definitely do so. I think it'll become more common with Conroe. I've got an MCW30 myself that's waiting to be put to use. I bet plenty will gain overclocks with Conroe if they step up the cooling on the northbridge.
I agree, I'm planning on going to Conroe toward the end of summer and it makes me glad I already have the MCW30 :thup:

[ECH]

Wow, I guess I am the only one that notices a difference. But I am using P660 not an AMD and I think because some do not undrstand Intel is why they think they are not needed. Currently I am using a fan on my NB and have been able to up my FSB by 10 so far when I couldn't before. I plan on WC it soon but they do not come with rubber or soft inserts to avoid metal/metal contact and to keep it from rocking about. So I have to fab something on my own.

[jkresh]

If watercooling is making much of a difference on the p4's/conroes why would motherboard makers put passive sinks on their high end boards, if they have the option of 450 fsb or 500 and they need active cooling to reach that it would seem silly not to include it? Or am I looking for logic in something driven by marketing (ie passive looks cooler and is quieter)

[ƒÓÒl]

Don't forget that most mobo makers don't support overclocking officially, they just give you some nice settings in bios and a wink and a nudge.
They know that if something goes pop they won't warrant a darned thing, claiming you took their product out of it's designed limits.

[natewildes]

In my Pentium D 805 rig, the NB heatsink my ECS SLI board shipped with was small and loud, and my chip wouldn't go higher than 3.8GHz. When I was building a water-cooled computer for my uncle, I tested it on my system (just a NB block), and I was able to get 4.0GHz. The only problem was that the ninja was practically turning red, but that's what I get for ditching water :D As people have said, Intel will benefit, AMD will ~not, and Conroe will hug and kiss you with happiness.

[jkresh]

ordered a block from EKwaterblocks

[Elif Tymes]

Mmmmm, Nothing like the taste of a freshly baked Chipset block. Mmmmm.

[Captain Slug]

Mmmmm, Nothing like the taste of a freshly baked Chipset block. Mmmmm.
I don't think it would be a good idea to eat something called "dying bread". :beer:

[Moto7451]

Like ƒÓÒl, I'm using Gemini Cooling's closeout blocks for part of my next rig (part of the chiller system actually :) ) and my bro Legionare45 is using them in his system. If anyone wants a good deal on a chipset block, go over there and grab one. You're looking at a ~$10 investment for good quality hardware.

http://www.geminicool.com/orderChipset.html

[TheSonOfX51]

Moto7451, I really dont like the idea of mixing cheap parts and water inside a case holding 1500 dollars worth the electronics...

It just doesnt look trustworthy.

[Moto7451]

They're only that cheap because they're no longer making them. These blocks are contemporaries of the White Water. The company that makes them is a CNC machining shop and is focusing on that instead of making blocks. They're as well made as any of the blocks out there for sale that cost $50 if not better in some cases (pointing to the initial Apogee blocks which had metal shaving issues).

I use their products in my systems and I worry more about the barbs on my brother's radiator leaking than I worry about his waterblock having problems. They seal their blocks with either O Rings or RTV silicone (on the lexan topped blocks) and I haven't had one leak problem with them.

[Silversinksam]

It all boils down to cost. The mobo manufacturers will only spend what is needed so it will work properly.
There is also a segment of market share geared to enthusiasts like us, there they will spend a bit more so the mobo has some headroom, but they charge for this amenity like a fan on a chipset :beer: