Test report Apogee GTZ vs. Koolance 350 vs. HeatKiller 3.0

Printable View

Show 100 post(s) from this thread on one page

05-22-2009, 10:51 AM
Freaky Freezer

Quote:

Originally Posted by santiagodraco

No. If Gabe wanted to be called out for being a scam artist, a liar and a fraud, then he'd follow your suggestion. Or don't you remember the Belkin fiasco? But I have not nor do I intend to suggest he is even close to any of those. He's a CEO doing his job to promote his product and NOT promote the competitors (which he would quite possibly be doing if his tests were broader in scope). But that doesn't mean he shouldn't be called out on it. He's a big boy, he can take the heat. And if he can't he shouldn't be CEO in the first place.

As for the figures, question 1 was already answered, and you can go HERE to see them. But to answer your question 1 again, 78.7% run at high speed/flow and 18.1 run at low or medium speed.

As to question 2, of those with one loop only 13 run CPU only and 86% run CPU+GPU. So that part of the test is more representative than the flow portion. However of those with 2 loops 64.3 have an isolated CPU loop and the other 36% have CPU + soemthingelse in one of the loops.

So, again, I think Gabe's results are accurate. I don't think he cooked the books or falsified any of the results. But that doesn't mean the conditions are not favorable to his block, and I would expect any CEO who publishes a test of his companies products to make damn sure that the tests ARE favorable or they DON"T get published.

But do so in a public forum and you better damn well expect it to be scrutinized.

Oh, and one other thing. Gabe is complimented left and right for doing his job. What the hell are you talking about. And as for implying that all other tests are invalid because of YOUR conspiracy theory is bull:banana::banana::banana::banana: as well. You attempt to credit Gabe by discrediting everyone else simply because they have the good sense to not post their own tests on a public forum and instead do so on their websites where it belongs. Here's what I would suggest. Gabe can create a Swiftec forum and posts tests to his hearts content. I would respect him for doing so. If you think the tests are valid you can start a post saying "Swiftec publishes new test results for medium flow loops" and I'd bet the response here would be quite different. Then Gabe can continue to do the other side of his job which he does VERY well which is to help his customers with questions about his products. I respect him and the others (like Dean from Koolance) for doing the same. THAT is the proper use of the forums by a company representative imo.

I'm not sticking up for Gabe. I don't know him on any form of personal level. I'm sticking up for his results, which you have admitted are an accurate representation of the majority of water-coolers on this forum. And I never discredited anyone else. I gave you plausible scenarios which you are obviously to obtuse to consider and would immeadiately call them "conspiracy theories".

You have no idea who the other independent testers on this site are, you have no idea who they work for, what business, company they are a part of or own. You have no idea what company gives them free stuff or work in exchange for "favors". YOU are the one who discredited Gabe because YOU DO KNOW WHO HE IS. If you're going to do that, then you need to descredit the others for NOT KNOWING WHO THEY ARE. Either take everyones results with the same level of trust, or don't trust any results, don't pick and choose.

Here's a conspiracy theoryfor you; you like Koolance and these results don't like Koolance best.
05-22-2009, 11:53 AM
NaeKuh

Quote:

Originally Posted by Freaky Freezer

Here's a conspiracy theoryfor you; you like Koolance and these results don't like Koolance best.

Heres another one.

Skinnee's test show both the KL-345 (but DAYAM HORRIBLE MOUNTS) and the KL-350 in favor of the GTZ on his bed.

So what does that tell ya?

To me its tells me its about application and how you set it up.

Anyhow Gabe generously gave out blocks to our testers, and they been testing it. Skinnee is almost done with the block roundup and ive been leaked special information (more like rants and the usual boredom crys on Load waiting), while he's been doing the test.

So Gabe doing these tests for us is a bonus, give the man some credit, and you guys he's very smart as well.
05-22-2009, 01:03 PM
santiagodraco
Quote:

Originally Posted by Freaky Freezer

I'm not sticking up for Gabe. I don't know him on any form of personal level. I'm sticking up for his results, which you have admitted are an accurate representation of the majority of water-coolers on this forum. And I never discredited anyone else. I gave you plausible scenarios which you are obviously to obtuse to consider and would immeadiately call them "conspiracy theories".

You have no idea who the other independent testers on this site are, you have no idea who they work for, what business, company they are a part of or own. You have no idea what company gives them free stuff or work in exchange for "favors". YOU are the one who discredited Gabe because YOU DO KNOW WHO HE IS. If you're going to do that, then you need to descredit the others for NOT KNOWING WHO THEY ARE. Either take everyones results with the same level of trust, or don't trust any results, don't pick and choose.

Here's a conspiracy theoryfor you; you like Koolance and these results don't like Koolance best.

Your examples are conspiracy theories meant to discredit, and you know it ;)

Ok, maybe this is what you are asking for: "Any manufacturer that posts results under assumed names are frauds, and you know who you are!!!" Is that what you are saying I should agree with? Now, as reasonable (if dramatic) that statement is, how does that somehow imply that tests not conducted by some manufacturer's employee are invalid? Don't you see anything wrong with your statement? You are basically saying that we should essentially dismiss all other tests because of the possibility that they are ACTUALLY done by the manufacturers...and you don't think that is conspiracy theory? You think that Gabe's tests are to be treated as absolutely alturistic simply because his name is attached? Man, you have a lot to learn about reality.

And you don't seem to be able to read statistics well. Gabes tests are NOT representative of most of the users of the forum. When he runs his pumps at full, not half, speed, then they will be valid for the "majority" of users.

Yes, I think Gabe's tests are valid, for the conditions of the test. I said it before and I'll repeat it again. But I also think the test is intentionally constructed to favor the GTZ. There you go. When Gabe runs a test that uses:
- Multiple rates of flow
- Uses single AND multiblock configurations
- And posts all results for all conditions
I'll be happy to call king to the king, whoever that may be. I'll be happy to call the GTZ king under Gabe's conditions, but that's it, and since I run MY pump at max speed it's not king for me, at least not according to other tests.

As for me liking Koolance (and you liking Swiftech - translation hypocrite) I've never attempted to hide that. I've also never spoken out against other tests that were fair, even when Koolance came out behind, since in most cases the tests were broad in scope and were representative of the varying conditions XS readers will see in the real world. But of course that doesn't fit your argument so you chose to dismiss it out of hand.

Now here's why I am using a 350 and not a GTZ.
- First I use ALU, and because of that the plated 350 is a better choice.
- Second the 350's mounting system is less prone to problems (ala it will work on any cpu variant, lapped, unlapped, too high too low).
- Third I like Koolance and what they have done, and the block looks good.
Now, if I was using no ALU I might have considered a different block, maybe the GTZ, maybe something else. But I'm happy with Koolance and no one can dispute that their recent products are top notch. Are they the best? That depends on the conditions. Certainly not always and certainly sometimes.
05-22-2009, 01:06 PM
santiagodraco

Quote:

Originally Posted by NaeKuh

Heres another one.

Skinnee's test show both the KL-345 (but DAYAM HORRIBLE MOUNTS) and the KL-350 in favor of the GTZ on his bed.

So what does that tell ya?

To me its tells me its about application and how you set it up.

Anyhow Gabe generously gave out blocks to our testers, and they been testing it. Skinnee is almost done with the block roundup and ive been leaked special information (more like rants and the usual boredom crys on Load waiting), while he's been doing the test.

So Gabe doing these tests for us is a bonus, give the man some credit, and you guys he's very smart as well.

I completely agree NaeKuh, certainly the tests add value and Gabe deserves credit, and yes he's obviously intelligent as are many of the folks that both participate and read these forums.

Let's see how Skinnee's tests go. I am very excited to see them.
05-22-2009, 03:53 PM
gabe

Quote:

Originally Posted by Martinm210

I never noticed where there was some sort of travel distance limitation,.... Of coarse I don't lap my processors very much, only enough to remove the coating and level the surface over the core, so it's probably a very small thickness.

travel limitation: there is; you may not have observed it, but it's there all the same. I wouldn't mention it here otherwise :) A washer doesn't fix the problem. A free travel mount does. I verified it.
very small thickness: we all thought the same thing (that it shouldn't affect the contact patch that is), myself included, prior to verify it by testing, but it does. Obviously how much it affects the contact patch depends on how much you lap the IHS; but it doesn't take much I can vouch for that. I lapped my 920 no more than 5 minutes (hand pressure) on the lapping machine, and it made a huge difference in the contact patch (I lost good contact with stock mounting). I didn't measure exactly how much I removed, but from 10 years experience with my lapping machine I can tell you that I removed between 0.003 and no more than 0.005" (<0.12mm)
05-22-2009, 04:23 PM
gabe

Quote:

Originally Posted by santiagodraco

.. or attempted to discredit it.

If when I say that the vast majority of users do not lap their processor, and that tests performed on lapped processor do not represent fair test conditions, then so be it: I do take exception to that data. [edit and clarification: the data is reported as gathered, but it's just flawed, unbeknownst to the tester]

Incidentally, I do realize that this is no different than you saying that I conducted my tests in a limited environment that is not representative of the average user. I accept this criticism. The difference being that I warned users in my introduction that the test conditions were aimed at Xtreme members. What I take exception to is you misrepresenting my intentions. They were clearly stated.
05-22-2009, 04:28 PM
LOUISSSSS

well i think all points have been argued in this thread, the take the results the way you want them. heres how i take it: the only difference between my setup and his is the extra pump, which (i think) wouldn't make much of a difference in temps if i added on to my loop. so i can say his results would be the same as mine. hence, i'm happy with my gtz, beating out the KL350 and also getting very close w/ the more expensive HK3.0
05-22-2009, 04:51 PM
Martinm210

Quote:

Originally Posted by gabe

travel limitation: there is; you may not have observed it, but it's there all the same. I wouldn't mention it here otherwise :) A washer doesn't fix the problem. A free travel mount does. I verified it.
very small thickness: we all thought the same thing (that it shouldn't affect the contact patch that is), myself included, prior to verify it by testing, but it does. Obviously how much it affects the contact patch depends on how much you lap the IHS; but it doesn't take much I can vouch for that. I lapped my 920 no more than 5 minutes (hand pressure) on the lapping machine, and it made a huge difference in the contact patch (I lost good contact with stock mounting). I didn't measure exactly how much I removed, but from 10 years experience with my lapping machine I can tell you that I removed between 0.003 and no more than 0.005" (<0.12mm)

Interesting...I just thought a travel limitation would only occur if the hold down plate was resting on the retaining e-clip which I didn't notice. Also visual inspection of the TIM coverage in block removal always looked good (or at least as good as other blocks), but not having a non IHS spread to compare to, it's pretty hard to tell visually.

Not the first nor the last time I'll miss something ...only way to know for sure would be to test it as you've done..:up:

Look forward to seeing some of this data, I'm also curious to see what lapping in general does for net benefits.
05-22-2009, 05:20 PM
liguhy

Quote:

Originally Posted by Sadasius

It's what I call 'the clap'. It's when they tighten it so much it folds the motherboard and claps them in the face...:rofl: ....er wait it might be something else.... I am so confused!:shrug:

sig quote! sig quote!
05-22-2009, 05:41 PM
gabe

Quote:

Originally Posted by ThaWaxShop69

Not true, a washer on the backplate will pull the backplate away from the board itself and make the heatsink push tighter on the chip (ive done this MANY times with IHS'less chips so i know it works)

I understand what you are saying, and it could have been a valid suggestion had the diameter of the screw body been smaller than the motherboard mounting hole (thus allowing the screw to 'penetrate' the mobo so to speak); unfortunately, this is not the case.

We are preparing a kit for lapped processors, using standard M3x35 screws as posts, and thumbnuts. It will require measuring the compression (as with other similar systems), but at least will allow adequate mating.
05-22-2009, 05:49 PM
gabe

Quote:

Originally Posted by Martinm210

Interesting...I just thought a travel limitation would only occur if the hold down plate was resting on the retaining e-clip which I didn't notice. Also visual inspection of the TIM coverage in block removal always looked good (or at least as good as other blocks), but not having a non IHS spread to compare to, it's pretty hard to tell visually.

Not the first nor the last time I'll miss something ...only way to know for sure would be to test it as you've done..:up:

Look forward to seeing some of this data, I'm also curious to see what lapping in general does for net benefits.

yes, to see the difference, you need to compare contact patches before and after lapping. it speaks loudly (at least to me).

Lapping improvements depend on the blocks. I am confident that independent testing will demonstrate that. I also measured it.
05-22-2009, 05:57 PM
antiacid

I can't help but laugh at the people getting their panties in a bunch over a .5degree difference on a waterblock given particular conditions. I mean, both the GTZ and the HK3 are great blocks but at the end of the day, ease of use will win the GTZ a lot more praises than anything else. If anything, Swiftech should be praised for the innovative mounting method that they introduced rather than on the marginal temperature differences that there may or may not be between those three top block :P
05-22-2009, 06:23 PM
nikhsub1

Quote:

Originally Posted by antiacid

I can't help but laugh at the people getting their panties in a bunch over a .5degree difference on a waterblock given particular conditions.

Don't take this the wrong way but, you will be laughing a lot. .5C is big around here.
05-22-2009, 06:43 PM
Sparda

Quote:

Originally Posted by nikhsub1

Don't take this the wrong way but, you will be laughing a lot. .5C is big around here.

0.5° is the average log data taken from a long interval right? If the value taken into normal observation isnt difference barely seen? :confused:
05-22-2009, 07:32 PM
Freaky Freezer

Quote:

Originally Posted by NaeKuh

Heres another one.

Skinnee's test show both the KL-345 (but DAYAM HORRIBLE MOUNTS) and the KL-350 in favor of the GTZ on his bed.

So what does that tell ya?

To me its tells me its about application and how you set it up.

Anyhow Gabe generously gave out blocks to our testers, and they been testing it. Skinnee is almost done with the block roundup and ive been leaked special information (more like rants and the usual boredom crys on Load waiting), while he's been doing the test.

So Gabe doing these tests for us is a bonus, give the man some credit, and you guys he's very smart as well.

I agree 100%

Quote:

Originally Posted by santiagodraco

Your examples are conspiracy theories meant to discredit, and you know it ;)

The only person my conspiracy theories are meant to discredit is you. You are the one who took Gabes results with a grain of salt, SPECIFICALLY stating that Gabe being the CEO was doing nothing but brand promotion with these tests. Your comments were insulting and completely inappropriate, accusing Gabe of doing nothing more than marketing and doing it by skewing results.

Your accusations had no substance other than you knowing that Gabe is the CEO of Swiftech because he makes that fact known to everyone on this forum. And if you're going to accuse him of such practices just because YOU KNOW he is the CEO of Swiftech, then accusing other members of doing the same thing because YOU DON'T KNOW who they are would be just as good of a reason.

I have a GTX on my current system. I will anxiously use my Fuzion v2 for my next build and I will use whatever is said to be best overall (price/performance/installation/support) at that time when I get around to building an i7 system. I have no loyalty to Swiftech or any other brand.

Quote:

Originally Posted by gabe

What I take exception to is you misrepresenting my intentions. They were clearly stated.

That was indeed the problem.
05-22-2009, 08:52 PM
santiagodraco

Quote:

Originally Posted by gabe

If when I say that the vast majority of users do not lap their processor, and that tests performed on lapped processor do not represent fair test conditions, then so be it: I do take exception to that data. [edit and clarification: the data is reported as gathered, but it's just flawed, unbeknownst to the tester]

Incidentally, I do realize that this is no different than you saying that I conducted my tests in a limited environment that is not representative of the average user. I accept this criticism. The difference being that I warned users in my introduction that the test conditions were aimed at Xtreme members. What I take exception to is you misrepresenting my intentions. They were clearly stated.

I stand behind what I say Gabe. And you said "aimed at Xtreme members" which would indicate that for that audience your tests are invalid since your testbed is not representative of "most" extreme users. Had you said they were "aimed at users who run lower flow loops with multiple blocks" it would be QUITE different. As it is the purpose of your test is to state that the GTZ block is the better block for the majority of XS users, which is untrue, based on your test conditions.
05-22-2009, 08:55 PM
LOUISSSSS

Quote:

Originally Posted by santiagodraco

I stand behind what I say Gabe. And you said "aimed at Xtreme members" which would indicate that for that audience your tests are invalid since your testbed is not representative of "most" extreme users. Had you said they were "aimed at users who run lower flow loops with multiple blocks" it would be QUITE different. As it is the purpose of your test is to state that the GTZ block is the better block for the majority of XS users, which is untrue, based on your test conditions.

so how would YOU propose the loop be set up to suit most members of XS?
05-22-2009, 09:04 PM
santiagodraco

Quote:

Originally Posted by antiacid

I can't help but laugh at the people getting their panties in a bunch over a .5degree difference on a waterblock given particular conditions. I mean, both the GTZ and the HK3 are great blocks but at the end of the day, ease of use will win the GTZ a lot more praises than anything else. If anything, Swiftech should be praised for the innovative mounting method that they introduced rather than on the marginal temperature differences that there may or may not be between those three top block :P

I'd bet there are a lot more folks reading XS that are NOT very experienced with watercooling. Because of this they will tend to look at the tests and pick the block on top (in my opinion). So if that is true wouldn't you hope that the test will be representative of what most of those readers do with their setups?

As for Swiftech's mounting method I'm not sure how it's so innovative considering. It's also limited to certain sockets without kits and is essentially a takeoff on the design of certain other block hold down plates (with less socket flexibility) from what I can see.
05-22-2009, 09:08 PM
santiagodraco

Quote:

Originally Posted by LOUISSSSS

so how would YOU propose the loop be set up to suit most members of XS?

I think the common denominator that seems to be missing is testing with these conditions:

1. Varying flow rates from off the shelf pumps.
2. Single radiator and single pump configuation.

The use of multiple blocks is certainly true to how most XS users are configured, including myself, so I'm not sure that needs to be changed. I guess purists would suggest that a product be tested alone, so It might be nice to see the same tests with and without a GPU block (since there's something like 30ish% of XS users that run that way).

If you are interested here's the link to the current survey results that show how 62 XS readers have their WC setups configured: Survey Results XS WC Users
05-22-2009, 09:09 PM
Waterlogged

I'm waiting for Richard Cheney to make an appearance in this thread. :rolleyes:
05-22-2009, 09:15 PM
santiagodraco

Quote:

Originally Posted by Freaky Freezer

I agree 100%

The only person my conspiracy theories are meant to discredit is you. You are the one who took Gabes results with a grain of salt, SPECIFICALLY stating that Gabe being the CEO was doing nothing but brand promotion with these tests. Your comments were insulting and completely inappropriate, accusing Gabe of doing nothing more than marketing and doing it by skewing results.

Your accusations had no substance other than you knowing that Gabe is the CEO of Swiftech because he makes that fact known to everyone on this forum. And if you're going to accuse him of such practices just because YOU KNOW he is the CEO of Swiftech, then accusing other members of doing the same thing because YOU DON'T KNOW who they are would be just as good of a reason.

I have a GTX on my current system. I will anxiously use my Fuzion v2 for my next build and I will use whatever is said to be best overall (price/performance/installation/support) at that time when I get around to building an i7 system. I have no loyalty to Swiftech or any other brand.

That was indeed the problem.

We'll have to agree to disagree then I guess :)
05-22-2009, 10:14 PM
shazza

Quote:

Originally Posted by santiagodraco

I'd bet there are a lot more folks reading XS that are NOT very experienced with watercooling. Because of this they will tend to look at the tests and pick the block on top (in my opinion). So if that is true wouldn't you hope that the test will be representative of what most of those readers do with their setups?

As for Swiftech's mounting method I'm not sure how it's so innovative considering. It's also limited to certain sockets without kits and is essentially a takeoff on the design of certain other block hold down plates (with less socket flexibility) from what I can see.

Agree with your first sentence. Have you tried the GTZ? It is extremely friendly to the first time water cooler, and would be a good choice for many people.

You keep saying you have no agenda against Swiftech, but you have certainly spent a lot of time trying to counter the results presented here ...

Show 100 post(s) from this thread on one page