Informal test results CPU 350 vs Apogee GTZ

[eXa]

you guys are idiots, you realize that if the gtz gets 1.86 from the ddc3.2 then the 350 is gonna be down to ~.86 gpm and its just gonna perform worse

Your point is valid, your math is not. What if you reduce the gpm on the gtz to 1 , cant just shave of 1gpm. The flow rate would probably be just above 1gpm for the 350.

[SNiiPE_DoGG]

Your point is valid, your math is not. What if you reduce the gpm on the gtz to 1 , cant just shave of 1gpm. The flow rate would probably just above 1gpm for the 350.

well, that's what the ~ was for but I would be willing to wager that it is <1 because the ek supreme can easily see sub 1 gpm flowrates

[[XC] riptide]

This thread is making me embarrassed and I find myself wincing as I read. *sigh*. Can you guys not understand that if he uses teh same pump in the dayum tests as he used everything else the same.... that it leaves only the blocks as the variables???? <<< Hence the whole idea of the test? I also note some snide remarks AFTER admin folk came in and warned.

Anyway, heres the bottom line : IN a 'real world' enviroment with 'retail' pumps we'll see a temp diff in favour of the GTZ, and I can say this purely on what Gabe has told us even not knowing the pump he used, as long as I know that he used the same pump and the same settings for voltage.

More tests will ensue I'm sure from the general community.

[T_Flight]

This thread is making me embarrassed and I find myself wincing as I read. *sigh*. Can you guys not understand that if he uses teh same pump in the dayum tests as he used everything else the same.... that it leaves only the blocks as the variables???? <<< Hence the whole idea of the test? I also note some snide remarks AFTER admin folk came in and warned.

Anyway, heres the bottom line : IN a 'real world' enviroment with 'retail' pumps we'll see a temp diff in favour of the GTZ, and I can say this purely on what Gabe has told us even not knowing the pump he used, as long as I know that he used the same pump and the same settings for voltage.

More tests will ensue I'm sure from the general community.

That's pretty much the way I understood it too. You can't test blocks unless you isolate the variables and that is exactly what Gabe did. I read some of the posts, but couldn't figure out what they were trying to get at because this testing is as good as it gets. Even Nik said he has *been there* and seen the test setup and said it was an excellent test rig.

I am very grateful for this data. This is the kind of thing that is extremely valuable. This kind of Data is the bottom line. Everything else is just guesses without data like this. I know i couldn;t even begin to run these kinds of tests. I don;t have the equipment for it. I am very grateful to the people who do though and the ones who spend their time to help people like me.

I wanna personally thank Gabe for providing these test results.

[Archmage]

It's true that I cannot figure any reason to discredit the results of this test.

However: As consumers and as rational people it is prudent to question results and remain skeptical, especially if those results contradict with what we've previously been told.

What variables could there be? Honestly...I can't find any that make enough sense to request a re-test.

-Higher heat load? That might be the most valid request because we ALL run overclocked CPUs. But still...why would the result be so different?

- Lower coolant temperature? Many of us run ~30C coolant temp so it was a standard choice. However in my uncontrolled testing, I've noticed some variance in deltas with different water temps. -- I don't have precise control like swiftech does, however. 25C and 35C perhaps? - meh...

- Flow rates? That wasn't much of a question for me. He used a powerful pump setup obviously, and that's enough. It seems like he maximized the performance of both blocks and eliminated the pump heat variable (possibly).

- This would depend on the position of the pump and where the water temp. measurement was taken - however it should not account for the magnitude of difference in performance measured anyway.

- Finally, to restate my last post, the pump's position in the line may make a difference. I think HondaCity had better temps with his 350 using the pump directly before the block.

I am defending my previous post. I do understand enough to grasp this thread, however there are some SMALL variables to address. I'm still satisfied with what GABE provided, however. I'm still a fan of the GTZ, and I'll consider it if I upgrade. Additionally: from back in the Procooling BillA days... I've always liked Swiftech

[migueld]

Anyway, heres the bottom line : IN a 'real world' enviroment with 'retail' pumps we'll see a temp diff in favour of the GTZ, and I can say this purely on what Gabe has told us even not knowing the pump he used, as long as I know that he used the same pump and the same settings for voltage.

More tests will ensue I'm sure from the general community.

Well there has been testing already in a more "real-world" context, that suggests a temp diff in favour of the 350. I'm referring to Hondacity's and NaeKuh's tests. Those tests showed an edge of 1-2C over EK's supreme. So IMO, we need to be careful about drawing conclusions in either case.

[hellcamino]

Well there has been testing already in a more "real-world" context, that suggests a temp diff in favour of the 350. I'm referring to Hondacity's and NaeKuh's tests. Those tests showed an edge of 1-2C over EK's supreme. So IMO, we need to be careful about drawing conclusions in either case.

I thought that is what I remembered as well, if the Ek Supreme beats the GTZ (in Martin's tests) and the koolance block has been seen to edge out the Supreme...It would seem logical to think that the koolance block would be the winner here although it also looks to take the award for most restrictive block as well.

[knightwolf654]

i really dont understand whats so hard in understanding the data gabe has provided, he set his pumps to a precise amount of GPM and using a chiller to keep the water temp at 30C taking out any possible ambient temp problems.
and ran both blocks. both the same flow rate and the 350 has a lower GPM because its more restrictive. what is the point at running both blocks at the same GPM when in the real world a pump is set at a set single set flow rate and it be impossible to run the 350 at the same GPM as the GTZ since the 350 would need a stronger pump. so in order for gabe to get the same GPM he would have to crank up that pump of his to probable higher then a real world pump could to get that 350's gpm up to 2.33 like the GTZ, but what happens when you run the GTZ at the new flow rate that took the 350 to get to get to 2.33? guess what is raises too.

what can you guys not understand here. dang man go back to school.

[septim]

Riptide, its just some can't read and apply the data given...

Ranker, not enough toes have been stepped on yet... Good Entrance, nothing less expected of you.

[DementeD]

i just dont understand how someone so new to the world of water cooling can sit there and try to argue professionals or people that have been on here doing this longer then they have been using a computer to play solitaire..just amazes me..

anyway great testing...

and i do believe thats its ok to question things and try to figure out stuff..but seriously when u have over half of the guru's on here pretty much saying your an idiot ..its time to sit back, watch and learn..

[nikhsub1]

I thought that is what I remembered as well, if the Ek Supreme beats the GTZ (in Martin's tests) and the koolance block has been seen to edge out the Supreme...It would seem logical to think that the koolance block would be the winner here although it also looks to take the award for most restrictive block as well.

Remember that Martin tested on a quad core which was 2 separate dies. Blocks can perform different on a center single die CPU like the i7.

[Bojamijams]

Heh, as if gabe is going to comment back on this thread.

Why should he waste his time explaining himself to people who lack the simple skills to understand that the GPM measurement simply refers to the resistance the block placed upon the loop.

5 pages later, and people still can't grasp this simple fact.

[NaeKuh]

Yep, your right, I was too broad on that statement. Modern I guess was meaning pretty darn new.

I retract my statement as meaning it's way the best. Still way better than the 350. All depends on your overall loop.

I got caught up in the moment. And thats a bad thing here. Data darnit, data!

and what happens when you add more pressure to a starved block which gabe's tests shows the 350 as?



You guys seriously need to learn how to read gabe's data...

Seriously... then your allowed to fight and complain..

Until you guys learn that... just say thank you gabe, and move on..

[situman]

WTF exactly. If this was a test of which block was more restrictive, then what you are saying might make sense. But of course you are not trying to make sense, you are trying to confuse the subject and win points.

The test is which block outperforms the other block in a real world environment. A proper test would be to determine which block performs best under various conditions, so that the consumer can make a choice that best fits the conditions they will use the product under.

Sorry if that makes more sense than you are comfortable with.

How many people can simple turn up their pumps to accommodate a more restrictive block? How many pumps can someone add just to compensate for a more restrictive block? This test is to show, at a given constant pumping power, how both blocks perform. If you turn up the power of the pump for one block, you must do so for the other block as well to keep testing constant.

[DeanB]

I personally would like the know the flow rate and static head pressure of the pump before either block is in place. These are test results I would like to have a hand at myself .

[Zehnsucht]

It's manufacturer showdown time

Koolance in this thread pitting against Swiftech, EK in another swinging at Aquacomputers

Nice, we like healthy competition of this kind!

[Waterlogged]

It's manufacturer showdown time

Koolance in this thread pitting against Swiftech, EK in another swinging at Aquacomputers

Nice, we like healthy competition of this kind!

http://www.youtube.com/watch?v=WvufFwdqMzg

[DeanB]

Lol pitted against eachother? I'd prefer not that kind of terminology . Competition is healthy

These are the tests that help us improve our blocks. What I want is verification of the results and identification of all the variables and their values. Bowed blocks and micro-grid arrays can be pretty complex, as both Gabe and Stephan know, in terms of "best mount, lowest temperatures".

[SiGfever]

What is most important in my opinion is that the top waterblock manufacturer's are constantly striving to improve their blocks. One day one is on top and the next day another. This is healthy competition and will advance the design and performance so that the end users are the winners.

[Stealth42o]

It still comes down to this,

If you don't like or believe the results Gabe provided based on his testing methodology, then set up your own bench and run your own tests. It's not that hard.

andyc

I would in a heartbeat if I didn't have these pesky bills to pay... and a child to raise... and a wife to.. attempt to keep happy

I miss Martin's Testing.

[Zehnsucht]

Lol pitted against eachother? I'd prefer not that kind of terminology . Competition is healthy

I'm sorry if I offended you, it was certainly not my intention. What I mean is when you ask for numbers like that, you want to do similar measurements on your own, because you want to defend your product It's nice, then we can see both sides doing measurements.

[migueld]

Nice, we like healthy competition of this kind!

[voigts]

I would in a heartbeat if I didn't have these pesky bills to pay... and a child to raise... and a wife to.. attempt to keep happy

I miss Martin's Testing.

I would like to get into testing too since it seems quite interesting. However, aside from the money issue, or should I say lack of money issue, I have no place to set up. I usually end up tearing apart my rig whenever I upgrade on the dining room table as it is (depending on the outside temperature in my workshop-small kerosene heater only), so I don't think my wife would be too happy fighting for a corner to eat on while the table stays full of watercooling stuff.

I am very grateful for this data. This is the kind of thing that is extremely valuable. This kind of Data is the bottom line. Everything else is just guesses without data like this. I know i couldn;t even begin to run these kinds of tests. I don;t have the equipment for it. I am very grateful to the people who do though and the ones who spend their time to help people like me.

I wanna personally thank Gabe for providing these test results.

+1. Well said.

[NaeKuh]

I miss Martin's Testing.

you guys know skinnee read this thread and went OH BANANAS!!!

u guys are scaring all our testers.

and give up on martin... seriously, the guy did more then enough for us.

Its gonna be a while b4 he gets bored again. And when he does, he knows i have a boat load of stuff for him to play with.

[ranker]

you guys know skinnee read this thread and went OH BANANAS!!!

u guys are scaring all our testers.

and give up on martin... seriously, the guy did more then enough for us.

Its gonna be a while b4 he gets bored again. And when he does, he knows i have a boat load of stuff for him to play with.

I think what most of us miss will be his unbiased testing. There are very few "straightshooters" on these boards when it comes to testing. Quite frankly, I won't trust anyone other than Nikhsub1, Petra, Martin (in that order), when it comes to testing. Too many people have their own "preferences" on these boards when it comes towards testing. It may not be apparent, but by reading their posts, you can notice the bias.

It's not easy to do these tests. I appreciate anyone that does them. However, it'll take time and a lot of unbiased testing before many of us will elevate a tester's results to command the respect of the three I mentioned above. I think that's why so many of us miss the testing that the three above do. Losing one of those three is a big loss to the community. =P

Hopefully Skinnee can adopt a similar methodology and have a similar unbiased approach.