Anyone seen the D-Tek Fuzion yet??

[STEvil]

Maxxx, CoreTemp and TAT both read from on-die sensors.

Not necessarily perfect, but repeatable dT.

[cdelong]

Well I got confirmation today from D-Tek that my Fuzion shipped out. Now the wait begins.

All good things are worth the wait.

[Petra]

All good things are worth the wait.

Provided that everything ships on time, we should have the Fuzion blocks by early next week

[h20 frag-monger]

I completed the 3 coldplate shootout on my test bench at work, glad my boss is out this week

A little about the setup: I used a 30x30mm heat source hooked up to a DC power supply outputting 100 watts, a Lytron chiller was used to keep the flow at a constant temp and to regulate flow, a load cell was used to make sure each block was mounted to the same pressure, an inline flow meter measured flow rate and a pressure transducer on the inlet and outlet measured pressure drop across the coldplate, temp was monitored by thermalcouples at the center of "case" of the heater and also in the fluid flow at the inlet and outlet.

-Thermal resistance was calculated by {(Tcase - T inlet fluid) /power}
-Pressure drop was simply {inlet pressure - outlet pressure}
-After a 2 hr block burn in i waited for steady state and took data at .5gpm, 1gpm,1.5gpm and 2gpm










As you can see the Fuzion has better thermal performance at all flow rates but what is more impressive is the pressure drop!

with pressure drop this low this block will be perfect for systems that have alot of components (like multiple water cooled cpus / AMD 4x4) and you could run them off one pump.....i need to make a performance vs pump power graph ....in my experience thermal performance increase as pressure drop increases (a-la storm) but the ability of the Fuzion to have this good of a performance with this lillte pressure drop is incredible.

one other thing to note is that my chiller's pump could not push more than 2.5 gpm through the Storm, and 3.5 gpm with the apogee, but the fuzion would go all the way to 4 gpm (close the max of my chiller) and its performance kept getting better where as the thermal performance of the other two were leveling off.

[jtdigital]

nice work h20 frag monger

[serialk11r]

But if it's kinking then they're not actually getting a tighter bend.


But it doesn't.

And BunBun threw a fit when I said that the turbulence inside 1/2" tubing was greater than the turbulence inside the Storm waterblock...until he proved it to himself.

In this area, the things people have deduced using common sense have proven to be more wrong than right. That is why I try hard to make sure that everything I say has some reasonably trustworthy data to back it up. I make sure of that because I know that common sense isn't good enough.

LOL yes, turbulence in tubing greater than turbulence in Storm. Hmmmm... I have to think about that one. If it is kinking it isn't getting a tighter bend true. But if it doesn't kink it is getting a tighter bend, don't you think? My masterkleer is deforming and kinks when I try to put it over barbs(Because I use my twisting strategy, not too good for tubes). But I don't mind, because I won't have sharp bends. And one thing, edit what you said about "I trust manufacturer specs". Because its wrong. Masterkleer is supposed to have a bend radius of 1. some inches, but when you bend it that tight, it kinks and the rest of the tube starts to collapse on itself.

[serialk11r]

OMG. Cathar needs to see this, storm got pwned...

[SiGfever]

This looks very impressive so far. It would be nice to see where the MP-05 SP LE fits in the equation.

[Sparky]

Wow nice work H20!

If the block really performs that much better than anything else in real life on a real CPU I just have to get myself one of those for my watercooling setup (whenever I can afford one )

[Graystar]

OMG. Cathar needs to see this, storm got pwned...

Ummm....someone please correct me if miscalculated...but aren't we talking about 0.2C at 100 watts? That's hardly "pwned"...

Of course, that doesn't account for pressure drop...guess someone will figure that one out.

[nikhsub1]

Ummm....someone please correct me if miscalculated...but aren't we talking about 0.2C at 100 watts? That's hardly "pwned"...

Was thinking the same thing...

[Graystar]

I completed the 3 coldplate shootout on my test bench at work, glad my boss is out this week

Impressive work. Congrats.

[serialk11r]

Ummm....someone please correct me if miscalculated...but aren't we talking about 0.2C at 100 watts? That's hardly "pwned"...

Of course, that doesn't account for pressure drop...guess someone will figure that one out.

0.2C at 100W, but with less pressure drop too. You said it yourself. Whats so not pwned? Higher flow with a teeny bit better performance is what I call pwn. It beat the storm in all aspects.

[h20 frag-monger]

Ummm....someone please correct me if miscalculated...but aren't we talking about 0.2C at 100 watts? That's hardly "pwned"...

Of course, that doesn't account for pressure drop...guess someone will figure that one out.

Graystar, your correct about the .2, but the heat flux of my 30x30mm heat source is much lower than that of a C2D...so results are gonna be closer.....in a real world test with the higher heat flux of the C2D i was seeing a couple degrees or so.

[nikhsub1]

Very nice h20 frag-monger! Got any shots of the heating element?

[Graystar]

0.2C at 100W, but with less pressure drop too. You said it yourself. Whats so not pwned? Higher flow with a teeny bit better performance is what I call pwn. It beat the storm in all aspects.

Well, you know, there were these claims of 5 C(or F) better temps floating around and other hoopla. If you consider that a “pwning” performance...as I’ve said before, everyone is entitled to his opinion. Personally, I'm keeping my Apogee (well, it's a 1U so it's not like I can change it anyways! )

[STEvil]

frag-monger, got a 6002 handy?

Now i'm even more tempted to mod my 6002 to have 4 outlets like I was going to do a long long time ago... lol.

[Graystar]

Graystar, your correct about the .2, but the heat flux of my 30x30mm heat source is much lower than that of a C2D...so results are gonna be closer.....in a real world test with the higher heat flux of the C2D i was seeing a couple degrees or so.

Well okay that's impressive.

Thanks again for the quality work!

[Petra]

Ummm....someone please correct me if miscalculated...but aren't we talking about 0.2C at 100 watts? That's hardly "pwned"...

Of course, that doesn't account for pressure drop...guess someone will figure that one out.

Looking at that test, while the performance gain over the Storm (temperature-wise) isn't anything to write home about, doing so at less than half the pressure drop of the original Apogee is nothing short of amazing.

I'm afraid I may have passed judgement on this block a little too quickly... However, I'm left wondering how things would stack-up on a smaller die-sim (or, perhaps, a live C2D modded TTV-style...guess I'd better get to work ).

[flou]

Good jobs guys really nice!
H2o are you going to make any test in cpu's?THanks!

[serialk11r]

Well, you know, there were these claims of 5 C(or F) better temps floating around and other hoopla. If you consider that a “pwning” performance...as I’ve said before, everyone is entitled to his opinion. Personally, I'm keeping my Apogee (well, it's a 1U so it's not like I can change it anyways! )

This computer stuff just has so many factors, that claims could just be the user's own fault. What if the person who got 5C better temps happened to put too much thermal paste and not apply enough pressure? There are just so many possible outcomes that could come to be as a result of this kind of stuff, you need a whole lot of 5C claims to say it really does get 5C better, because the chance that everyone is screwing up and saying BS is much smaller.

[BillA]

always a pleasure to see 'professional' testing, nice h f-m
do not neglect consideration of the relatively 'huge' heat source, not so sraight forward as it seems

in any case Swiftech has shot themselves in the foot, again
sigh, 1.6M shares of stock not worth jack sh*t

[nikhsub1]

in any case Swiftech has shot themselves in the foot, again
sigh, 1.6M shares of stock not worth jack sh*t

ROFL! Sorry Bill, but that was funny

[flou]

This computer stuff just has so many factors, that claims could just be the user's own fault. What if the person who got 5C better temps happened to put too much thermal paste and not apply enough pressure? There are just so many possible outcomes that could come to be as a result of this kind of stuff, you need a whole lot of 5C claims to say it really does get 5C better, because the chance that everyone is screwing up and saying BS is much smaller.

C'mon, did you read who made the test?Some of the best overclockers,the whole idea was hosted by Swiftech.The guys told everyone,bring one block that can beat our new block, Apogee gt,and if you win, you will get a gift. I don't think that these guys,some of the best overclockers and people from swiftech, would like to say such things/or they made mistakes on the setup!

[h20 frag-monger]

Very nice h20 frag-monger! Got any shots of the heating element?

yeah this is it, its got cartridge heaters with an embedded TC and the housing is made of G10, we tried to make it as isothermal as possible (except for the top surface of coarse)