Swiftech Apogee GT vs the D-Tek FuZion CPU blocks @ CES

[Xavior]

Have you ever actually handled a pure copper bar before? It is far from soft. Its soft compared to other metals, but is not what I would call soft. The bases of those blocks there are rigid enough alright.

have you ever thought of how soft copper is, and the great amounts of pressure the mounting mechanisms put on the block?

Relatively soft, compared to other metals. WSP isn't talking about soft as in ''teddybear soft'', obviously.

Even though a copper base isn't exactly ''soft'', it's going to bend (slightly) under the stress of the mounting mechanism.

[clokker]

Even though a copper base isn't exactly ''soft'', it's going to bend (slightly) under the stress of the mounting mechanism.

I didn't think any of this discussion involved base warping due to mounting pressure (which I should think would lead to a convex base anyway), rather it was assembly pressure at issue (the larger diameter o-ring creating a slight convex distortion as the block block itself is screwed together).

Didn't Cathar have a few instances of this with the G5 due to a slightly shallow o-ring channel (or maybe a slightly too large o-ring...can't remember which)?

[WeStSiDePLaYa]

Have you ever actually handled a pure copper bar before? It is far from soft. Its soft compared to other metals, but is not what I would call soft. The bases of those blocks there are rigid enough alright. The thing is, if you look at the thermal conductivity of AS5 or any thermal paste, you will find that it suxxors and might as well be your poop(well not really). If the distance between the block and IHS or core increases by just a tiny bit, the thermal paste will have to "bridge" it and hence the crappy heat transfer and the crappy temperatures.

that is exactly my point

and my little cd/glass thing was just to emphasize the theory for those who werent understanding that strong mounting pressure will reduce contact over the core.

and that the bowed base helps counter that occurence.

so we are on the same page, just different paragraph.

[Mr. Peanut]

no offense taken just curious as which of the many statements I made there I'll be proved wrong on? I did a lot of fast typing and assumed everyone was inside my head thinking some of the same things as me.

By stating whatever yeilded the most surface area would be king, I also figured it was obvious I meant surface area contact..... and yes pressure plays a role in that as well

I was referring to your statement about surface area>all. I took that to mean that a flat IHS combined with a flat block would equal optimum temps. However, looking at the results at CES, this was not the case.*** I find it hard to believe that a bowed base will become perfectly flat under mounting pressure. (Although I invite someone to prove me wrong on that point.) Therefore, wouldn't a bowed base yield less contact area between the base and the IHS than a flat base?

I should have qualified that statement to pertain only to IHS topped CPU's Of course eliminating a "barrier" to the heat transfer process (the IHS) would obviously yield better results.

An IHS isn't the only barrier between the base of the block and the die. There is also thermal paste. If contact between the base of the block and the IHS is improved, then the "barrier" of thermal paste would be reduced. Under your reasoning, this would decrease temps, which we have seen happen. I think that the bowed base's performance has just as much to do with good contact as it does with anything else.

***I would also like to emphasize my belief that everything anyone, including myself, posts about bowed bases is purely speculation at this point. More testing needs to be done with them and the factors that may or may not help them to gain or lose an edge in performance.

[Plywood99]

Couldn't all this stuff be done away with by simply making the blocks perfectly flat, and using a different mounting mechanism?
One that would apply pressure over the center of the block:

___V___ - mount
======= - block
_++++_ - cpu

Seems just about all current mounting mechanisms force some sort of warping of the base...

[nikhsub1]

Well I can tell you guys one thing, if you hard mount a FLAT apogee GT with too much pressure you will bugger the base for good - I have just managed to do so.

[virtualrain]

Isn't this result far more likely than distorting the copper base when you apply too much mounting pressure on a block?



Even this is probably more likely depending on the rigidity of the mounting plate...



I suppose if you really go overboard, you may start to distort the base itself but I would think the major distortion would occur on the mounting plate long before the base got distorted.

Nikhsub1... what did you do to your base?

[WeStSiDePLaYa]

Isn't this result far more likely than distorting the copper base when you apply too much mounting pressure on a block?



I suppose if you really go overboard, you may start to distort the base itself but I would think the major distortion would occur on the mounting plate long before the base got distorted.

Nikhsub1... what did you do to your base?

id say the mounting brackets are easily harder than the copper bases. they would have to be to work properly.

and if a mounting plate was made like that, the base would still deform at high pressure.

[nikhsub1]

Nikhsub1... what did you do to your base?

http://www.xtremesystems.org/forums/...d.php?t=130318

[virtualrain]

id say the mounting brackets are easily harder than the copper bases. they would have to be to work properly.

and if a mounting plate was made like that, the base would still deform at high pressure.

I guess it really does depend on the rigidity of the mounting plate... it also depends on how much the intermediate layer (not shown in my diagrams) will give as well... usually blocks are like...

Mounting Plate
--------------------
Delrin or IMP Housing
--------------------
Copper Base

So the rigidity of all three layers comes into play as well as the torque on the screws making it difficult to determine just how much the copper is going to distort, if any.

[cdelong]

I was referring to your statement about surface area>all. I took that to mean that a flat IHS combined with a flat block would equal optimum temps. However, looking at the results at CES, this was not the case.*** I find it hard to believe that a bowed base will become perfectly flat under mounting pressure. (Although I invite someone to prove me wrong on that point.) Therefore, wouldn't a bowed base yield less contact area between the base and the IHS than a flat base?

An IHS isn't the only barrier between the base of the block and the die. There is also thermal paste. If contact between the base of the block and the IHS is improved, then the "barrier" of thermal paste would be reduced. Under your reasoning, this would decrease temps, which we have seen happen. I think that the bowed base's performance has just as much to do with good contact as it does with anything else.

***I would also like to emphasize my belief that everything anyone, including myself, posts about bowed bases is purely speculation at this point. More testing needs to be done with them and the factors that may or may not help them to gain or lose an edge in performance.

Without restating everything you just have- I think we're basically on the same page here, but I'm making assumptions you know what I'm talking about without stating every explicit detail Yes the stock goop between the IHS and the die blows.... that being said, I'm not sure we can reasonably or safely put enough pressure on the IHS by cranking down on the WB base to make Intel's goo perform any better under the IHS. I just think we just may be stuck with what's there... and every CPU is as different as you and I in that respect. I don't know... maybe the installation force does squeeze the TIM under the IHS??

I just spent the last 1.5 hours lapping my E6600. Very funny profile Highest at the edges (caused by solder joints??), next highest in the center (die and TIM??) and lowest in between. It only took 5 minutes to see that the edges were the highest. It took the other hour and change to see the rest for sure. Lapped down to 1500grit and burning in now. I'll let it go for a few days and formally post what I see. Unformally, I like it so far (this is with the Fuzion).

I'm going to have to get a GT after my PA120.2 is set-up and get busy. I've asked D-Tek for a new midplate to fuss with and should have one soon. If they do implement a nozzle I'm all game to give it a try. Hopefully it will be a simple midplate upgrade or just an insert to install.

[gabe]

Very funny profile Highest at the edges (caused by solder joints??), next highest in the center (die and TIM??) and lowest in between.

exactly what we saw at CES with Fugger's quads.. This is apparently a result of the soldering process.. HOWEVER, I just checked an AMD IHS today that we had removed from an X2 4600, and it had the same wavy profile (high edges, high center, low mid-section), but to a lesser extent than the Intel.

[Curtybob]

I might be mistaken, but I thought the IHS was soldered to the die? Yeah... I remember seeing an article about a guy trying to remove a couple, and wound up with some cool shots of the dies removed from the packaging.

[neonflx]

please see my comment here

http://www.xtremesystems.org/forums/...86#post1965986

i could not finish orthos before because of high temps now i can

[pumbertot]

as i dont have an apogee GT im finding it hard to visualize where this gasket(that has 2 thicknesses) actually goes in the assembly. can someone explain it? can we do anything similar to a storm or any other block or is only the apogee built in such a way as to warp it?

[situman]

so which is the preferred choice? Fuzion or GT?

[Turtle 1]

I am sorry but if you look at the D-Tek casting . mechanicly speaking the mateing surface and the mounting surface are on a parrallel but not the same plane. that casting has a vertical wall that meets the mating surface. Where the metering block mounts to the base is a differant plane. Its much more rigged than the apogge block . I can see where installing a larger diameter o-ring would cause deflection of its mating surface.With the apogge.
The same does not apply to the D-tek design . I highly suspect that the convex surface of the D-tek block is the result of the casting process. Fact is I believe the D-tek block when it comes out of its mold is actually concave . and is mechanicly straightened to a convex with a press. The reason I believe this is the mass of the pins . In the cooling process would cause warp towards the mass.

Looks like you should have paid more attention to this post. You may than have avoided ruining a perfectly good block.

[darcon]

so which is the preferred choice? Fuzion or GT?

Seriously.
And in a simple, straight-forward answer, not one filled with a bunch of useless theories and bias.

[nikhsub1]

Seriously.
And in a simple, straight-forward answer, not one filled with a bunch of useless theories and bias.

*Sigh* - there is no straight forward answer here... there are lots of variables and not one block will be best for all. It also depends on what CPU you are running. I will have more results in a couple of weeks.

[redcorn]

Check out this cheap effective way to protect our motherboards using intel 775 chips and any water cooling cpu block.

http://www.xtremesystems.org/forums/...=1#post1972879

[nealh]

Check out this cheap effective way to protect our motherboards using intel 775 chips and any water cooling cpu block.

http://www.xtremesystems.org/forums/...=1#post1972879

Gabe...Dtek...you need to provide a backplate for the mobo with these blocks

[Levish]

Gabe...Dtek...you need to provide a backplate for the mobo with these blocks

If the springs are used (and the instructions followed) you wouldn't get nearly as much warping as pictured, even less once you mount the mobo in a case since the motherboard hold downs would aid in keeping it from bending.

[epion2985]

So has anyone put a pelt under either one of these yet? Interested in how well they would perform next to the mcw6002. The base seems large enough to accommodate a 50mm-50mm pelt and the cooling patch is large. Any word yet?

[redcorn]

If the springs are used (and the instructions followed) you wouldn't get nearly as much warping as pictured, even less once you mount the mobo in a case since the motherboard hold downs would aid in keeping it from bending.

I barely had my Dtek tension screws turned 5 turns each when I took that picture. Anyone using these water blocks should definitely take the time and spend 5 bucks to install a back plate to protect your mother board. Just for fun I went back and installed the Koolance 300 V series cpu block and when tightened to factory specs the motherboard bows. What I think is happening is people are installing the blocks while the mother board is in the case so you never realize the board is bowing. Install the block on a bench or table and you will see it happen fairly easy.

[Philly_Boy]

So has anyone put a pelt under either one of these yet? Interested in how well they would perform next to the mcw6002. The base seems large enough to accommodate a 50mm-50mm pelt and the cooling patch is large. Any word yet?

Haven't gotten there yet in my testing. . .maybe next week. . .

I am interested myself to see if the heat dump from the pelts affect the CPU overclock. We'll know in awhile.