Recent water block (GT/Fuzion/Storm) test results...

[dacooltech]

In response to someone earlier in this thread, the reason no one talks about the MP-05 anymore is that it's no longer in production. It's discontinued and has been completely replaced by the FuZion.

They are not even made by the same company. Are you sure the Mp-05 is discontinued, I still see them on the cooltechnica site?

Methylphenidate, Fairydust is absolutely right and you are wrong. AquaXtreme MP-05 Pro and SP Limited Edition blocks are still in production and available through select resellers such as Petra's Tech Shop, Jab-Tech, Crazy PC etc.

The only water blocks that we discontinued are the White Water and White Water LE blocks.

D-tek was the primary producer of the MP-05. I've spoken with Danny personally about MP-05 availability. He is sure no one is still producing them.

CoolTechnica is/was the sole owner and manufacturer of the AquaXtreme MP-05 CPU water blocks. Like I said we are still manufacturing MP-05 PRO and SP LE blocks and as long as there is demand for them we have no plans otherwise.

Bruce
CoolTechnica / AquaXtreme

[D-TEK]

To straighten out some misinformation....The MP-05 and all the other AquaXtreme blocks were made with D-Tek money. Sorry but I will not sit here and see all that sole owner crap.

They may be in production still but there is no D-Tek money going into them anymore. As a matter of fact We are owed quite a bit from this ordeal and now have to chase it.

I suggest this subject ends

[Hassan]

[dacooltech]

To straighten out some misinformation....The MP-05 and all the other AquaXtreme blocks were made with D-Tek money. Sorry but I will not sit here and see all that sole owner crap.

They may be in production still but there is no D-Tek money going into them anymore. As a matter of fact We are owed quite a bit from this ordeal and now have to chase it.

I suggest this subject ends

Danny,

Apparently you're the one who's been misinforming the customers that D-Tek was manufacturing the MP-05 blocks and discontinued the production and all that crap.

It's also not true that MP-05 and all of our AquaXtreme blocks are made with your money... and you know it...
Let's say you let X company in on a deal and they start offering your Fuzion block under X brand... now does this make X company own the design and manufacturing rights of your Fuzion block?
unless you sell the design and manufacturing rights of your product you're still the sole owner of Fuzion right?

CoolTechnica is the sole owner of MP-05 water-blocks and they're still in production. period!

I suggest you to read the e-mails that I sent you betw. October, and early December last year - that you did not even care to respond -. Also check the attachments...

If you want to start a flame-war that's fine... But my past experience taught me that there are no winners engaging in a flame war situation.

I suggest you to deal with me directly to straighten out any unresolved issues that you may have. I've been dealing with a family emergency and hopefully be available after tomorrow. You have my numbers...

Bruce
CoolTechnica / AquaXtreme

[Jupiler]

Danny and Bruce,

Although we appreciate the input and support within our forums, I don't think this is the place to start this kind of discussions between 2 companies.
There are other ways to discuss this other than here at XS.

Thanks for understanding.

[D-TEK]

Bruce,

So I sell a block with your name printed all over and I claim that we manufactured it? We discontinued the blocks that we paid for development and manufacturing on and did NOT receive!!!

This block would NOT have been made without us. If you want to hang with your story to protect your image go right ahead. I have the documented proof of the situation along with the transfer receipts all in the hands of a firm right now (you also know that).

BTW.. an email is on the way. I want to see what your attachments are and see what can be done to settle this.

Jupiter,

I am sorry.. This will be my last post on the subject as long as Bruce does not feel the need to keep this in the public eye. I will move it to email. Be free to delete this entire spat if you like. His dilusions required a response.

[nikhsub1]

To get this back on track, I finally have my test rig ready to be assembled. I have many blocks to test - I do NOT have an MP-05 LE though, if Bruce wants to send me one to test I will gladly include it. I hope to have the test rig running by the end of this week.

OT - Bruce, I wish you and you family the best - I also know what it is like to have a child with a medical condition so my heart goes out to you.

[SiGfever]

@Nik,

This is good news about your test station. I have been waiting for someone to test the MP-05 SP LE against these other blocks. I have an MP-05 and feel it is the finest of my blocks and have been very happy with its performance. I recently installed the FuZion and find it to be a very high quality block also and with less restriction it has a lot going for it. The MP-05 has a large surface area and performed as well as the FuZion on my current Presler p4d 945 @4.5Ghz.

I am looking forward to your results. And by the way, now ya got me thinking about a PA160 rad for a possible second loop for a NB and SB on either a P5W DH or an D945XBX2 since I have my DDC-2 replacement and might just try a little smaller tubing for them say 3/8"id.

[babalouj]

I see that not many people have compared the fuzion and the mcw6002a. I just upgraded from my 6002a to a fuzion in preparation for quad core. My setup is as follows:

p5b deluxe
e6600 9x400 @ 1.47V w/ lapped IHS
mcp655 @ max setting
mcr320 w/ 3 yate loon SLs
mcw6002a and Dtek Fuzion
thermalright 775 backplate
Arctic Silver Ceramique
Coretemp for temp monitoring and Orthos blend for load, load temps were taken after 15 minutes.

My results were kinda suprising. These are the best numbers after 3 remounts:

mcw6002a
ambient: 22C
idle core 1: 33C
idle core 2: 35C
load core 1: 51C
load core 2: 51C

Fuzion
ambient: 22C
idle core 1: 33C
idle core 2: 36C
load core 1: 51C
load core 2: 51C

Im sure the gap widens in favor of the Fuzion on quad core but until I buy a quad core I can only test the dual core performance. For all of you dual core users that have a mcw6002a, you wont see much/any improvement with the fuzion apparently. This really isnt the thread to be posting this but I didnt want to start a whole new thread just to compare another waterblock to the fuzion.

[virtualrain]

Added Kunaak's latest results to the summary page.

Also added a link to your post above (Babalouj) for those interested.

I'm looking forward to adding some GTX test results to this thread... Get at it boys!

[Kunaak]

I too want to see some GTX testing.
looks like a real interesting block indeed.

[Fossil]

What all the tests seem to show is that your mount matters more than your choice of block. Maybe block manufacturers need to focus a lot more on mounting, particularly getting good mounts on dodgy IHS surfaces.

Most tests are done with lapped CPUs and blocks, and still the mounting variation is large - and a lot of people are not going to want to go anywhere near lapping their IHS.

With the Swiftech GTX allegedly producing a difference of 1 or 2 C, you are probably better off trying remounts until you get a good temp, rather than going out and buying one to replace a GT, or even an old Apogee, quad-core quibbles not withstanding.

With a restricted loop, it would seem that the FuZion is a no-brainer, but again the mount seems a major factor.

I wish Intel and AMD would pay some attention to this and put some research into it. I am always hearing from people running stock coolers that are running extremely hot, yet my experience of stock coolers is that they are nowhere near that bad unless you are overclocking (just noisy). Surely, the mount is to blame, and making those affected machines practically unusable?
Mounts are an issue for everyone

[nealh]

Hmm..keep debating on whether a fuzion is worth replacing my Storm rev2......

[nikhsub1]

Hmm..keep debating on whether a fuzion is worth replacing my Storm rev2......

On the cpu in your sig? Definitely. I have started testing again - I have 3 mounts of the Storm and 3 mounts of the Fuzion done. Bill Adams has (and will continue to) look over my data. We decided to scrap the 24v run of the RD-30 as it just shows nothing useful - all temps just rise the same due to pump heat, ambient rises too. So I will be performing 5 mounts per block and all run at 18v. My cpu is UNLAPPED and no water block has been touched by me.

When Cathar talked about being at the theoretical limit of cooling he meant without an IHS. The IHS creates a whole new ballgame.

[dinos22]

excellent thread

it's tough sorting through all that info to find the best block

it seems that if you apply the bowed base mod to Apogee it slightly etches in front but the issue is that such mod benefits quad core CPUs mainly and no so much dual cores even thought they have a 1-2C improvement according to Gabe from memory


I think it's also pertinent to mention that these results are valid for specific types of processors as Storm used to be king until C2D and quads rolled into town and different waterblock designs were once again more efficient




interesting result with that MCW6002 block...............i was always fond of those blocks i must say

[dinos22]

When Cathar talked about being at the theoretical limit of cooling he meant without an IHS. The IHS creates a whole new ballgame.

yeah that's right.................today i realised that intel is also making CPUs with easily removable IHSs and not the epoxy method so it will be interesting if they continue that as Storm will once again be the best block if that happens unless the die size becomes too big

but even with IHSs onCPUs last year with AMD Storm wasn't taking a beating until intel

that's why i thought it should be mentioned that it is only since the new generation Intel chips that we have seen these changes


also another interesting development is occuring in air cooling. OCZ is moving towards a carbon/graphite based air cooler base which is supposed to have some amazing thermal properties (will have to wait and see about this claim) but it would be interesting to know whether same material could be used in waterblock designed

here are the two relevant threads on this material and cooling technique

http://www.xtremesystems.org/forums/...d.php?t=137493

http://www.xtremesystems.org/forums/...d.php?t=138253

[Cathar]

Okay, just want to carlify the finer points of my comments about the theoretical, but rather, the practical limitions of water-cooling, and what my comments were really about.

I was referring to the effective areal convection co-efficient given practical levels of pumping power and tubing as befits installation in a computer case. The comments still apply when IHS's are involved, but the differences seen between specific block implementations (i.e. Swiftech Storm vs some other blocks) is more of a matter of optimising for a set of scenarios, rather than invalidating the comments I made.

When we stick a waterblock on some heat source (CPU) the heat spreads through the copper of the waterblock base and is convected into the water flowing through the block. The heat spread through the copper can be calculated, and is termed as the "spreading resistance". Through measurement and some mathematics and some measure of estimation, we can define a waterblock's effective "cooling power" per unit of area that it is trying to cool in terms of watts per area per Kelvin, or W/m²K. We arrive at a value, typically between 10,000 and 120,000, and this is a measure of the effective water-metal convectional transfer rate, being an estimative mapping of the area that is being cooled (heat source contact area) onto water-cooling structure internal to the block's design.

I took to the time to measure and track the area cooling power of most of my block designs, as well as other designs, and it is really this that formed the basis upon how I determined whether one design was moving in the right direction. It is a unit of waterblock performance measurement that is independent of things like base irregularities, cooling patch size, base-plate thicknesses, IHS's, even the material out of which the waterblock is made out of. All those things are separate issues which can be controlled or optimised independently based upon the scenario in question. Ultimately, no block will perform well if its effective areal cooling power is crap, however it is very possible to have a block with a very high effective areal cooling power perform poorly if those variables are not a good match for the specific scenario at hand. This is precisely why we can observe designs with fantastic convectional efficiency perform worse than mediocre designs.

Warped pressure-sensitive IHS's on oddly dimensioned and/or multi-die cores was never in the design considerations of the Swiftech Storm, and it is for this reason that the specific implementation that was on sale doesn't perform as well as the design's convectional efficiency would suggest. It doesn't mean that the internal design is bad, it just means that all those external variables I mentioned above are significantly less than optimal for that scenario.

[Okda]

first time to see this wonderful thread

thanks for the good job

Kunaak (Fuzion vs Storm)
CPU: X3210 (quad core) OC to 3.6GHz@ 1.48V Vcore
WC Gear: MCP-655, Old BIX Extreme(?)
Test SW: Wprime
DT Results (Avg. from TAT): Fuzion = 42.9, Apogee GT = 52.8
Winner: Fuzion beats Storm by 9.9

the above quote needs to be corrected

[Cupcake]

Well for my naked X2 3800 @ 2.9 with 1.62v the storm rev1 beat my apogee by 8c

[gabe]

Well for my naked X2 3800 @ 2.9 with 1.62v the storm rev1 beat my apogee by 8c

Very likely because your apogee was sitting on the socket cambox and not making contact with the CPU. I just tested a naked 165 (Toledo) and noticed the Apogee was sitting slightly on the cambox. Ground the plastic a tad (1mm) with a dremel, and load temps now sitting at 38C~40C at 2.7 & 1.6v

[Fossil]

The issue with Storm always seems to have been not poor performance, or that it required a strong pump:
- From what we've seen in test after test, it still scores well (though not best) on dual and quad cores.
- A bad mount can make more difference than having a Storm or a FuZion
- The storm is fine with the pumps that everyone uses now

The Storm's problem is that it's restrictive and doesn't play as well in a multi-block loop as some other blocks.

The FuZion's main win is that it's very unrestrictive, and gains extra advantage in multi-block loops that are increasingly common (maybe).

Another issue with the Storm now is that it's getting a bit rare and expensive, as dealers aren't restocking it, and the price of blocks seems to be falling slightly while the Storm remains at an old price-point.

I was tempted to get one for my HTPC project, but the price put me off. I'll use the Apogee GT that's currently on my main machine when I rebuild ... which will be soon: FuZion and PAs arrived today!

So a question: are multi-block loops increasingly common? It seems logical now that GFX cards are becoming the main heat generator in a case that people will need to w/c that more than they did in the past.

I think with the CPU manufacturers still die-shrinking, and increasingly performance-per-watt driven as they are now, that we could start to see far more dual loops and even GFX only loops (with CPU on air) - and not for o/c but just to shut up those awful GFX card blowers.

[virtualrain]

the above quote needs to be corrected

Fixed! Thanks!

[virtualrain]

What all the tests seem to show is that your mount matters more than your choice of block. Maybe block manufacturers need to focus a lot more on mounting, particularly getting good mounts on dodgy IHS surfaces.

I agree... this was my conclusion as well.

What amazes me is that current water block solutions for Intel motherboards do NOT include any kind of backplate to stop the motherboard from warping or cracking under mounting pressure. How can you possibly get a good mount without a proper backing plate!?!

[serialk11r]

The issue with Storm always seems to have been not poor performance, or that it required a strong pump:
- From what we've seen in test after test, it still scores well (though not best) on dual and quad cores.
- A bad mount can make more difference than having a Storm or a FuZion
- The storm is fine with the pumps that everyone uses now

The Storm's problem is that it's restrictive and doesn't play as well in a multi-block loop as some other blocks.

The FuZion's main win is that it's very unrestrictive, and gains extra advantage in multi-block loops that are increasingly common (maybe).

Another issue with the Storm now is that it's getting a bit rare and expensive, as dealers aren't restocking it, and the price of blocks seems to be falling slightly while the Storm remains at an old price-point.

I was tempted to get one for my HTPC project, but the price put me off. I'll use the Apogee GT that's currently on my main machine when I rebuild ... which will be soon: FuZion and PAs arrived today!

So a question: are multi-block loops increasingly common? It seems logical now that GFX cards are becoming the main heat generator in a case that people will need to w/c that more than they did in the past.

I think with the CPU manufacturers still die-shrinking, and increasingly performance-per-watt driven as they are now, that we could start to see far more dual loops and even GFX only loops (with CPU on air) - and not for o/c but just to shut up those awful GFX card blowers.

I don't understand what you mean by the Storm being expensive. On jab-tech, its only 55 bucks...

[gabe]

Originally Posted by Virtualrain:
What amazes me is that current water block solutions for Intel motherboards do NOT include any kind of backplate to stop the motherboard from warping or cracking under mounting pressure. How can you possibly get a good mount without a proper backing plate!?!

This is the answer: Companies such as Swiftech that respect/follow the thermal and mechanical design guidelines established by the chip makers do not HAVE to use a back plate simply because the clamping pressure of our cooling devices to the CPU are on specs and therefore do not pose a threat to the motherboard integrity.

We have observed a number of products available on the market right now that cause significant warpage to the area directly underneath the socket. This is due to the fact that these products use extra stiff springs, and do not include a safety mechanism in the mounting system to prevent users from clamping the devices above specs, and I will add that this is a common cause to motherboard instability and/or complete failure. One example I could use was the dramatic mishap that happened to us at the CES challenge which caused our EVGA board to die prematurely. This only happened because for convenience reasons (multiple mounts of various water-blocks) we did not use our stock mounting mechanism –which again IS safe.

The above being said, we have been conducting extensive research and testing since CES, and we do recognize the desirability of a back plate as a means to enhance the clamping pressure safely. As a result, we do have a back plate in production right now, and will hopefully be able to release it by the end of April. It will be included with all existing and upcoming swiftech water-blocks. I would like to take this opportunity to ask the members of this forum to appreciate the fact that this project was not as simple as it looks from the outside. In effect, since there is no specification for a back-plate in Intel's 775 motherboards, manufacturers have been free to place all kind of surface mount components at the back of their motherboards… and they obviously don’t put them at the same place! So if you use the wrong back plate you risk crushing the SMC’s and destroying your motherboard in the process. A good back plate then is one that will be compatible with the majority of the popular motherboards (read "overclockers" models) available today.

In conclusion all I can say at this time is that we do the best we can with the tools at hand, and hope that motherboard manufacturers will keep this particular problem in mind next time they release their newest and greatest!

A final word to all Apogee GT and GTX owners: when the back plate becomes available we will also offer it for free to all who will ask (proof of purchase required) for a period of 30 days after release of the product.