• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Hoot's WB Shootout is on the Front Page

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
The gemini is a cool w/b and designed very similar to fix's spiral. I have seen one good incoclusive result in a comparison with the maze2 vs. fix's spiral, where the maze2 had the advantage with a small margin lower temp. I just don't see how the gemini is kicking so much booty. Now I have had instances where dismounting and remounting my w/b's has aided and hurt the overall full load temps by as much as 5c. So having the maze2 seated not so perfect might be a plausible explanation.

I know hoot is right on up there and also keeps note of surroundings and other pertenant data to insure his findings to be legit, or as close to right as you can, but I still find hard to believe the gemini w/b outperforms the maze2 by such a margin.

r12
 
Hoot is pretty concientous (sp) so I am sure that he did it correctly and followed the manufacturers instructions as well he should have in a shoot out. He did however remount it several times in an attempt to rectify what he thought were erroneous results to no avail. I have experienced results somewhat the same though I was not doing any comparison testing. Now, the results that I have achieved with the block in the past were dependant on removing the springs and hand tightening the nuts in a cross pattern. I do this again in a week or two to insure that the block is fully seated. Hoot did hint at this as he noted that "strangely" if he moved the block some the temps would lower. For the purposes of comparison testing he could not in good faith modify the mounting methodology of the manufacturer.
 
Hey Hoot, I really liked the article. Though the Gemini blocks ruled by a couple degrees, I like the mass of the Swiftech because it looks cool, and it still has massvie cooling power. I like higher powered pumps too.
 
Be leery of some "alleged" high power pumps. I know there are some out there, but many pumps claim astronomical flow rates and don't deliver. I just got a message from a fellow who had an "alleged" 340gph pump, but he did the gallon jug fill test and it was only delivering 72gph. The Eheims seem to deliver as promised and they're almost dead silent. I know mine does. You get what you pay for.

Hoot
 
I am aware of this. This is one of the reasons I bought an Eheim, also for it's quietness and reliability. Cost a bunch, but I think it's worth it. Next time I think I'll get a 1250, and tune the flow for temps.
 
racecar12 said:
The gemini is a cool w/b and designed very similar to fix's spiral. I have seen one good incoclusive result in a comparison with the maze2 vs. fix's spiral, where the maze2 had the advantage with a small margin lower temp. I just don't see how the gemini is kicking so much booty. Now I have had instances where dismounting and remounting my w/b's has aided and hurt the overall full load temps by as much as 5c. So having the maze2 seated not so perfect might be a plausible explanation.

I know hoot is right on up there and also keeps note of surroundings and other pertenant data to insure his findings to be legit, or as close to right as you can, but I still find hard to believe the gemini w/b outperforms the maze2 by such a margin.

r12

I had a hard time with this too but I know Hoot knows what he is doing. racecar12, are you refering to the posts at Athlon OC? Here is another thread that shows the Maze 2 besting the Spiral. Hoot's tests lend creedence to Peter's explaination of all the math homework he did on his block.
 
Yea. Peter's hard work has paid off. It has not only benefitted him, but also many in the Overclocking community. Now if there was only a way to convince people that water and computer do mix. :D
 
Hiya Hoot, I'm using the maze-2 myself and also found it to be a bit hard to get a good seating on the processor.

I just went ahead and tightened the nuts till the springs would go no lower. I experienced a 3 degree drop in cpu temp. I'm pretty sure a better radiator would've helped even more.

I was it might be cracked, but it wasn't. I'm not even using a shim!
 
adamtekh said:
so why no die simulator testing?

?????????

-and from a previous post in this thread-

adamtekh said:
so why no die simulator testing?

?????????

Adam, I know of two of these in existance. Do you have one I can borrow to test the Gemini blocks with. I don't have several hundred dollars to have one machined and to buy the power supply necessary to run it. I doubt Peter has one either. Harder yet would be getting my PC to run with it. PCs run with CPUs. I tried very hard, within my monetary limits to give all four of those blocks a level playing field to excel upon, using a common CPU. I know my XP1600 did not subject those blocks to the same challenge they would see using a die simulator, but I don't think there are a lot of readers out there just itching to strap one of these blocks to their die simulators. They are itching to strap one of these blocks to their common CPU.

I hope to get more blocks to test using the same setup I used in my article. Several manufacturers have already contacted me offering one of their units to test. Though I do not have MathCAD on my PC or a Degree on my wall, I will try not to disappoint them or the readers.

This small effort I put forth has spawned some kind of Techno-WitchHunt. If it were not for all the positive emails and PMs I have gotten the past day or so, I'd swear off doing anymore tests. Peter does not deserve any grief either. The poor guy simply designed and manufactured a series of water blocks that happen to work well with my setup. I suspect they will work well with other peoples setups also.

I'm going to go and melt some ice cubes with my forehead now.

Hoot
 
Last edited:
Great article Hoot! A big hats off to Peter for doing his homework and making a product that really performs. I particularly like the fact that he posts in here and included pics of the shop. Nice to see where the product is comming from, and the investment that went into getting the tools and equipment to manufacture the product. Good to see someone who actually takes the time to research (and most likely test endlessly) before he goes out and just puts out a mediocre waterblock. Once again my hat off to you Peter, and expect e-mail from me inquiring about the purchase of one of your water blocks in a few months. (gotta pay off some bills to free up my computer play cash...)
 
Yes CrystalMethod, I really appreciate the effort that Peter put forth in order to release such a great block. I also appreciate Hoots efforts in providing us with data that we can all look at to aid us in our decision making. Remember folks, Hoot didnt have to do this, he did this out of his own time to provide us with the data that many of us use to make our decisions for this and that. Just appreciate the work that other do in their quest to help fellow man, or in our case, fellow OCers.
 
Well Hoot, I thank-you for your efforts, and the article too.
I like the waterblock, and had since the site was first linked in here. It appears to me you did a fair job, and even tried to verify results that weren't expected....very nice job...hats off to ya!

As for the others, well...theory don't cool my cpu, copper and water do.:mad:
 
errr . . .

I recall seeing Peter's post indicating that he started with the theory

be cool
 
I think some of you guys rather flatter yourselves with regard to your percieved engineering skills. Of course there are efficiency gains to be made. Big ones.

Think about cars, back when every mechanically minded sob was building his own design car in the barn, say around 1905, we had 7 litre engines that did like 3mpg and 15mph, they mastered the basic concept, but not the ideas of thermodynamic efficiency.
Do some quick math on how far auto engineering has come in understanding this field. now currently the typical effiency overall of a car engine isn't all THAT good, but it's come a long way, lets say it's 80% efficient, that's generous, lets say 20% is wasted in heat, that's low, let's say we have a 100hp car, that's about 110Kw output, so let's say we have to design the rad and fan to remove 20Kw..... Now consider, an average vehicle rad, it's like 20x16inch, you get the same cooling area as 16 - 20 typical rad coolers that people might use on a PC, so that would suggest that you could like dump 1Kw or more heat through one of those, anybody even manage to dump anything close to that? We're still using circa 1905 mekkaniking skillz here people.

So we're like messing around with the equivalent of sleeve valve two lungers and Peter comes along with a smaller, lighter, 6 cylinder 4 stroke overhead cam engine, that he say's he designed thermodynamically rather than empirically, and it has 4 times the power and uses a quarter of the gas, and some of us are laughing at him, some of us are saying the gas is rigged, some of us are looking under the test stand for the other motor that must be necessary to get those dyno readings.......

While myself being somewhat more of an empirical spirit in the workshop, I am a physics BSc, so my empirical experience might be considered to have a little more refinement, anyway, I have been thinking about waterblocks since the mid 90s, on and off, and took a kind of back of the envelope stab at the theory occasionally. Anyway, when I first started looking into this stuff online when I got far more serious, I was pretty damn amazed at the pipe diameters and flow rates you guys were finding necessary. The phrase, sledgehammer to crack a walnut comes to mind.


Anyway, lighten up guys, whaddya wanna do, burn Armstrong at the stake for walking on the moon??

Road Warrior
 
Hoot said:


This small effort I put forth has spawned some kind of Techno-WitchHunt. If it were not for all the positive emails and PMs I have gotten the past day or so, I'd swear off doing anymore tests. Peter does not deserve any grief either. The poor guy simply designed and manufactured a series of water blocks that happen to work well with my setup. I suspect they will work well with other peoples setups also.


and

Originally posted by Road Warrior

Anyway, lighten up guys, whaddya wanna do, burn Armstrong at the stake for walking on the moon??

Let me jump in here for a moment and see if I can clarify some communication issues.

First, no one disputes that Peter has made a first rate water block. Whether he comes in above or below the Swifty and DD models, here's a guy who on his very first attempt is competing with the leaders of the pack. Peter deserves a lot of praise for bringing in a wb from scratch that gives the big boys a run for their money while being smaller and cheaper.

But there are some issues that do seem to be in dispute. First is the issue of whether Hoot's results can be extended to other systems, and hence whether the Gemini really is better than the swifty. Second is the issue of the theory behind Peter's block. I'll address these in order.

What I hope to acheive it the recognition that we can continue a substantial discussion of the issues in this thread, and even ask some very hard headed technical questions, without engaging in either a witch hunt, or burning Armstrong at the stake.

Hoot's testing

Hoot, being the careful, fair-minded tester that he is, never claims more than anecdotal strength for his results. He acheived those results by careful testing on a single system that he carefully describes. Your mileage may vary. Anyone who thinks that Hoot's testing doesn't provide good anecdotal evidence needs to go take a remedial course in research design.

Of course, we can still discuss just how extendible Hoot's results are. What would happen with different flow rates? WHat would happen with reverse flow? WHat about other blocks? WHat about systems with higher or lower heat outputs than Hoot's? These are all legitimate questions that are left open by Hoot's testing. (I think Hoot recognizes this.) The fact that these questions are left open doesn't mean that Hoot's testing is inadequate or flawed, it just means that it's anecdotal, which is of course all that Hoot ever claimed. Frankly, as anecdotal evidence goes, Hoot's is some of the best. Hopefully he'll consent to doing some more testing.

The distinction between the quality of Hoot's testing and it's extendibility is important. Questions about extendibility are not questions about quality. If this were not so, then almost every experiment done by the best scientists would have to count as poor quality since almost no experiments are perfectly extendible.

Peter's theory

I've already mentioned that Peter deserves Kudos. I believe that the engineers among us are happy to admit that. But in standard academic style they quickly pass on to the question of WHY the block works. Criticisms of Peter's theory do not amount to criticisms of his work. Peter has produced a first rate block. Whether he produced it by math, intuition, a vision from god, or reading chicken entrails, that fact remains. The engineers are trying to understand why the block works, not questioning whether it works.

But the question of why the block works is an open one. Petere can tell us why he thinks the block works, and we can debate whether his theory is correct. Doing so is not in itself a way of giving Peter grief. In fact having your theory questioned in this case should be taken as high compliment. It means you've produced something that deserves discussion. When an idea is just hogwash, it tends to be dismissed, not criticized.

It's important to understand the theory behind Peter's block. I mean the theory that explains why it works, not the theory that he used in building it. If we can understand the theory, that may help us to build even better blocks. If we ignore the theory, then better blocks will be a matter of sheer luck and probably won't happen. If we just take Peter's word for the theory, then if he's wrong we run the risk of not being able to create better blocks.

As good as Peter's block is, we want even better blocks. OUr best chance at getting them is to understand the principles that make Peter's block work. The best way to start on that is to ask Peter what theory he used in creating it, and then critically evaluate that theory to see if we can improve on it.

Academics and engineers can seem harsh when we start to criticize someone's theoretical stance. In part we're just so used to doing it, that we forget that other people aren't accustomed to it. Among ourselves, such criticism is often considered high praise. If I were in Peter's position, I'd be thrilled that people were taking me seriously enough to be highly critical of my ideas. But then, I'm an academic and used to that sort of thing.


So, I hope that the technical discussion can go on. The critics among us can try to remember that our ways can seem harsh to others, and the non-critics can try to remember that for some of us criticism is the highest form of flattery.

I apologize for the length. Some of you know that I am nothing if not long winded. But I hope that this can smooth some ruffled feathers and foster continued discussion of an important block and the theory behind it.

nihili
 
I believe it's important to note that Peter and Fixitt have resolved their differences. There is no need for anymore talk about who did it first.
 
I can't help but feel several of the remarks here have been directed towards me and my comments. I'd like to thank nihili for his words because they ring pretty true with me.

I have never meant to start a witch hunt nor attack anyone and I hope no one feels that way. But, I am a mechanical engineer with alot of experience in heat transfer and have spent years studying and analyzing ways to move heat. Because of this, I tend to ask the "why" and "how" types of questions. I realize that there is not a lot of real engineering knowledge out there in our little water cooling world, so I also try to help spread some when I can. I spend alot of time diving deep into the bowels of analytical analysis and design and love to crunch serious numbers.

However, I can't even begin to count the times that I've shared some proven engineering knowledge or questioned statements from others that do not understand heat transfer principles, only to be railed on and nailed to the wall by critics. Remember the big aluminum vs. copper heatsink debate and the article I wrote concerning it? Talk about a witch hunt! I've got e-mails that would make many shiver, got into more flamewars than I care to remember and was attacked as if I were Benedict Arnold. But now look; who doesn't believe copper makes a better heat sink?

Anyhow, to get back on subject...I think Peter has done a pretty good job on designing his waterblock, but I also think there is alot of room for improvement and tons of room for some real analytical analysis and calculations. That's why I asked Peter about his design philosophy and if he had done in numerical modeling. I'm not trying to put him down in any way. Same goes for Hoot; he did a great job with what he had in comparing the blocks. But if you read the other thread about why I think all numbers have to be taken with a grain of salt, you'll see why I don't just read a number and think it's perfectly perfect.

I love to see people working hard on our little hobby. But I also love to dig in and figure out why certain things work and how to better model designs and heat transfer so that better designs can be made. And finally, I also have a very, very hard time watching people post theories, equations and other assorted stuff that just isn't right. I'm sorry if I offend people by correcting them when they are wrong, but I'm not about to let mis-information continue to blossom across the internet.

And to those that are critical of engineers and our tendency to over-examine and over analyze everything, without that examining and analyzing, we'd never have put a man in space. Engineering principles are sound and proven. You only get one shot when launching something like the space shuttle for the first time, so you damned well better understand the physics and get your analysis right. Sure no lives are at state in water cooling, but that doesn't change the fact that proper analysis according to the physics of nature will lead to better designs.
 
nihili,
Very nicely put, and the work yet to be done is also well described. The path to improvement is through criticism, but few like their baby being called ugly. (old clichés)

At the start of this thread I complemented Hoot, and I suspect that his results will be found to be generally reproducible. Hoot’s results describe the wbs performance at one set of conditions, and the extendibility of the dataset will depend the accuracy and precision of the original data, and the ability to correlate new data to it.

As some are aware, I do (commercial bench) testing of watercooling components and have been nudging here and there to standardize the testing conditions to improve the comparability of test results. The identification of 25^C as the coolant test temperature (and a typical ‘good’ user value) is a first step, hopefully the same can be done for power and flow.

For those that may wish to comment, my suggestions are:
40.0, 70.0, 100.0, and 130.0 Watts, and
0.50, 1.00, 1.50, 2.00, and 3.00 gpm
(the described precision may not be attainable by all, just identify the variance)

Hoot has indicated he’ll forward the larger Gemini wb and, as I have a 462U, it should be possible to ‘back into’ Hoot’s test conditions by setting the flow to Hoot’s value and adjusting the applied power to reproduce the die temperature for two of his setups. However there are a number of iffy assumptions involved. And I will send Hoot my thermometer calibrator so the offset between our setups can be corrected for.

The on-going burden for any wishing to do testing involving temperatures will continue to be the accuracy of such measurements. This subject is not well understood by those without experience, but as the old saw goes: You can’t make a silk purse from a sow’s ear, or in techie terms: GIGO.

My questions about Peter’s theory have already been asked (and not answered). No biggie, not all questions merit an answer, I guess.

be cool
 
First of all, I thank those who answered my question. Now I'm satisfied. But it's sad that this has turned into somewhat of a flame war. Let's just get back onto the subject of this post.

On that note, who can recommend a good book for thermal physics? I'm in grade 12 so I don't want anything hardcore, but I did take General Physics and know the basics.

-DarkArctic
 
Back