Anyone seen the D-Tek Fuzion yet??

[nikhsub1]

I'm saying you didn't lap it flat, that would be like lapping a baseball, it doesn't work.

So the base of the FuZion is round? Like a baseball? Hmm, mine didn't look like that. Mine looked like this:

[IanY]

It is rumored to be slightly convex. I don't think it looks that way on first glance.

{edit} Your block does look absolutely flat, but eyes can be deceiving.

[Psychotic Emu]

Cheers niksub1, appreciate the effort!

Shame you have no storm but there should be a good correlation beween storm and G5 data.

I expect the dP over those is hugely different and given your pump setup I would suggest that in this situation the fuzion is probably getting minimal returns for all that extra flowrate it has over the G5.

Don't really know what my point is (bit tired, been working on finding a loss coeff of a fitting at work most of the day) but it may be something to discuss

[nikhsub1]

It is rumored to be slightly convex. I don't think it looks that way on first glance.

{edit} You're block does look absolutely flat, but eyes can be deceiving.

Yes I discovered that... but guess what? It is LESS convex when the base is unmounted from the top/mid plate. I believe the torquing of the screws plays a big role in why the base is convex...

[flou]

Just a moment, swiftech storm and storm G5 are different right?What is the difference among them?If anyone knows help!

[rosco]

In the first place, how certain are we that all C2Ds are concave, and are concave to the same degree? What if a consumer has an AMD with a flat IHS? That convex FuZion block surface continues to trouble me.

You can't be sure because IHS shape is higly dependant from the manufacturing process. I spent a lot of time to talk deeply about technics with the Intel manager who is responsible for thermal aspects, IHS, etc. IHS are very flat before the mounting and curing process on dies, but a mechanical stress appears and distorts IHS a bit or a lot, it depends. They can't lap them at the end because it's too risky, cost is higher and they lose the nickel plating too. Intel is aware about the "poor" quality of some IHS, but there isn't an easy way to correct it, but they try... This guy made some researchs about concave/convex and distorted bases to improve thermal transfer on some "bad" IHS too, I had the opportunity to see some of his internal publications, but it becomes complex to calculate and to manufacture the good shape. The easiest and best way for everyone is to lap IHS and WB flat yourself, but you loose warranty. It's your own choice.

Thx Nikh to bring some fresh info. I'm very curious about this bowed base of Dtek too. Did they confirm that "rumor" ? or is it the result of forging process ? or the result of the screws torquing if too tightened (thin base) ? Nikh's WB seems to have a quite flat reflection with no deformations. It's far more difficult to make a convex base than a flat one, so there's a trick here !

[Fr3ak]

Philly Billy,

You first post DTEK won fair and Square by 5 degrees. This test was done with DTEK in the room ....then you say your limo has arrived and you are flying home.....

Then magically you are back testing blocks again with only Swiftech employees and say they beet the Fuzion by 3 degrees. So it made up 8 degrees difference....How many shares of stock did you get!

Maybe you can read something between the lines that I cant. Here is what Philly_Boy posted:

OK, guys. . .I was at the Swiftech challenge last night. The limo is outside waiting to take us, but I will tell you in a word that the FuZion beat the Apogee GT fair and square by 5c. I was the one taking the temps and doing the averages. I will post detailed results along with many pics later today or tomorrow on the Swiftech Challenge thread.

Folks we have a new champ!! I will take this block and the Apogee GT home with me to compare it to the Storm and others.

That sounds to me like they were still on CES. He posted in a hurry the results and was driven to the hotel. There they did further testing.
Thats how it sounds to me.

Usually I would welcome new members, but in your case I will do an exception . If you dont like the result being posted, dont come to this forum or just dont reply.
I am sick of all the complaints. Without posting anything, it would have been a much more relaxed night for Philly_Boy. On the contrary he is even getting disrespect.

Sorry but that bad to be sad. I can stand a lotm but sometimes its enough.


Thank you nik for the fast results. The Fuzion seems to have a worse performance on dual core cpus with a smaller die size. I somehow expected that. Lets wait for more people testing the block to make final conclusions.
But from what has been posted, the Fuzion seems to be a low restrictive block, which is something I personally like a lot.

[Fameless]

Also, you say you lapped the fuzion. so right there all tests after that are voided when comparing the fuzion to anything else. as you changed the properties of the block, and how can anyone be certain you improved it?

If this is directed at nikhsub1, he did not lap the waterblock. He lapped the IHS of his proc.

[Graystar]

Yes I discovered that... but guess what? It is LESS convex when the base is unmounted from the top/mid plate. I believe the torquing of the screws plays a big role in why the base is convex...

That’s interesting. Maybe the delrin needs to be “faced”. In another life, when I was a bicycle mechanic we would “face” the surfaces of a bicycle frame where parallel bearing surfaces would be mounted. The two areas were the bottom bracket (crank shaft area) and the head tube (steering bearings). The mounting surfaces were separate but needed to be perfectly parallel. There are special metal cutters that perform the task.

With the Fuzion there are 4 faces that should be perfectly flat for a no-torque fit. I wonder if the hollow, thin-bottomed copper base is just too flimsy to withstand a not so perfect fit. Also, the deformation could simply be caused by the necessary compression of the gasket.

I guess the moral of the story is...assemble it, lap it, then don’t mess with it.

[flou]

We don't have also to forget that niksub uses 2 laing plus , maybe that thing boosts performance with storm, while with fusion it doesn't help in something a lot!

[h20 frag-monger]

wow...Nik i should send you my G4 and apogee ...cuz now im confused

i figured Fuzion would be par with the G5, since it had beat the G4 by 2 degree in my system test....

also, when you said you averaged the ambients, you mean you averaged the deltas between the ambient and the core temps over a ten min period right?

[nikhsub1]

The Fuzion base was lapped AFTER TESTING. So end that now please. frag-monger, I recorded the ambient temp every 10 minutes of each test... i then took the 6 ambient temps and averaged them.

[nikhsub1]

I just have been chatting with Stew on IRC... he rekons the G5 should outperform the G4 by a bigger margin on IHS cpu's... it has a much larger cooling patch than the G4. Perhaps it all does make sense now. BillA is sending me a Swiftech G4 to test... the picture (of my testing) will get clearer with more blocks run.

[h20 frag-monger]

Nik, that sounds like it should work, the only thing i can think of that would be a varaible is equal mounting pressure on each block?, other than that someone mentioned you were using 2 pumps, i only have one.....are your 2 pumps in serial or parallel....my guess is that they are in series (theoreticly doubling your pressure head) which should be benificial for the storm....if they are in parallel (thereicly doubling your flowrate) it would give the fuzion an advantage....you should see what the results are with one pump, 2 pumps in series and 2 pumps parallel....not that you need more testing to do, but it would be cool to see

where can we send all 5 blocks to get tested? stormG4, stormG5, apogee, apogee gt and fuzion

[h20 frag-monger]

I just have been chatting with Stew on IRC... he rekons the G5 should outperform the G4 by a bigger margin on IHS cpu's... it has a much larger cooling patch than the G4. Perhaps it all does make sense now. BillA is sending me a Swiftech G4 to test... the picture (of my testing) will get clearer with more blocks run.

cool, thanks billA, Nik do you want my apogee?

[Graystar]

I just have been chatting with Stew on IRC... he rekons the G5 should outperform the G4 by a bigger margin on IHS cpu's... it has a much larger cooling patch than the G4.

That's what I've been thinking all along. I can't see how some regular flow-over design is going to beat the tube-&-cup design.

[nikhsub1]

Nik, that sounds like it should work, the only thing i can think of that would be a varaible is equal mounting pressure on each block?, other than that someone mentioned you were using 2 pumps, i only have one.....are your 2 pumps in serial or parallel....my guess is that they are in series (theoreticly doubling your pressure head) which should be benificial for the storm....if they are in parallel (thereicly doubling your flowrate) it would give the fuzion an advantage....you should see what the results are with one pump, 2 pumps in series and 2 pumps parallel....not that you need more testing to do, but it would be cool to see

where can we send all 5 blocks to get tested? stormG4, stormG5, apogee, apogee gt and fuzion

Yes, I use equal mounting pressure! Want to know how? No springs! I tighten all nuts till I can't tighten them anymore. No lie. Yes my pumps are in series. Im going to run as many blocks as I can - I have a G4 and an apogee GT coming. What else is there? An MP-05 LE? If someone sends me one ill test that too. All the good testers with equipment like yours have moved on... your setup is the bomb! 30x30 is a bit big, I wish we could have like a 20x20 heat die with an IHS over it for testing, that would be ideal.

[h20 frag-monger]

i have got absolutely no real work done since i joined this forum

[nikhsub1]

cool, thanks billA, Nik do you want my apogee?

Got the apogee GT coming, dont need the old apogee Or do I?

[h20 frag-monger]

Got the apogee GT coming, dont need the old apogee Or do I?

i would just like to see if our results match up, but i guess we can do that with the stormG4 ...but if you want it, i'll send it

[afireinside]

Waiting on my mcr220 to come...

[Petra]

Now ots your turn nik and petra

I'm going... I'm going...

Slight delay, however, as the owner of the little machine shop next door took one look at what I need done, noticed Intel's "no fluids or oils" note, and told me that, while he probably could do it, he wasn't going to. He didn't want to risk breaking a bunch of bits and/or killing the CPU. Honestly, I really can't blame him.

Anyway, I'm going to check with the big shop that we've switched to for our tops tomorrow--they're far better equipped and seem to like challenges....

[Cathar]

Let's play "Spot the difference!"

The Storm design family:

Top Left: LR Storm/G7 Prototype
Top Right: LR Storm/G4 Final
Bottom Left: LR Storm/G5 Final (what nikhsub1 has)
Bottom Right: Swiftech Storm Rev 2

[flou]

So,what is the difference between the 2 bottom blocks?I suppose that you have run tests!

[cdelong]

Let's play "Spot the difference!"

The Storm design family:

Top Left: LR Storm/G7 Prototype
Top Right: LR Storm/G4 Final
Bottom Left: LR Storm/G5 Final (what nikhsub1 has)
Bottom Right: Swiftech Storm Rev 2

I can't measure your pics with a micrometer, but the two on the right look nearly identical.

On the left we seem to have a large difference of holes sizes, yet in the same basic pattern. My MP-05 SP LE midplate (MP kit nozzle #3) looks most like the bottom left pic- hole size and pattern.