Holy mega quote batman!
Quote:
Originally Posted by
UNOE 
Okay so I'm sure the best way to view this information is with the water temp. average into it. But will any of the core temperature average data be released on its own. This will help me to see where my mount is falling into this data. I can't measure water temps here but I can see the core temp.
Quote:
Originally Posted by
Agueybana_II
I will be getting the MIPS looks good and perform well.
Quote:
Originally Posted by
opt33
Just for my own curiosity, I averaged 9 different testers results from those that used equipment to measure accurately. I recorded for each block how many degrees C from best performing block in each of 9 tests. Then averaged the data for all 9 tests. If revision had changed for sure, tried to include newest revision.
In chart are the results in order of best performance when averaging all tests. With a few exceptions, not to different order from Strens results. And one note, a few sites have tested more than one sample of same block, and not uncommon to get 1+C difference between 2 samples of same block on same test setup, then add fact that different testers have unique cpu/IHS, and the fact that no block ever tests best in all reviews is easily understood. Also note alphacool nexxxus has only been tested twice, so really need more testing on that one, because all it takes is one less optimal sample and it would be right in middle of rest of pack.
Thanks for this it's always useful to know how consistent tests are. Although so much now depends on bow as so many of the blocks are so close.
Quote:
Originally Posted by
deafboy
Well I officially hate my block....great. lol.
Yeah the one definitive thing the test showed is that the Rasa is a very out of date design.
Quote:
Originally Posted by
opt33
I bought an alphacool nexxxos waterblock and was going to test and use myself if performed well in mine. Just to warn anyone else that is thinking of buying one, the out "threads" is only 50% threaded. See pic
here post 87. Mine is not going to be tested, but is going in the trash.
actually here is pic:
I'm going to attach a BP 40mm extended and see if I can get it to move at all. It's definitely a design flaw and gives you zero confidence but I wonder how much pressure you need to really make it leak.
Quote:
Originally Posted by
tsm106
Is that a defect or a design flaw? It's to tell from that pic.
It's definitely design mine is the exact same way. I would have found it when I did the internal pics - but I ran out of time before I left. Again big apologies for that
Quote:
Originally Posted by
UNOE
Okay so I'm sure the best way to view this information is with the water temp. average into it. But will any of the core temperature average data be released on its own. This will help me to see where my mount is falling into this data. I can't measure water temps here but I can see the core temp.
Yes but core temps are pretty meaningless without coolants temps, if you have ambients at least you have a chance, but even then there are so many differences with the radiators and varying flows that it's worthless for any real comparison.
Quote:
Originally Posted by
Agueybana_II
I will be getting the MIPS looks good and perform well.
The MIPS is a very good block, solid construction, fit, performance and flow.
Quote:
Originally Posted by
ugotd8
Occurs to me that stren is an experienced enough water cooler that he would have mentioned if his sample was half-threaded on the outlet. Must be a defect. I hav one coming tuesday, I'll let you guys know if it's design or defect for sure then.
Really sorry you guys went and bought these based on the review when they are flawed.
Quote:
Originally Posted by
opt33
I understand your skepticism, I am looking at it and have a hard time believing it,....but I will post some internal pics, you really have to see it to believe it.
I outlined the missing threads from arrow to dots then arrow. If threads were here at same depth all around, there will be restricted flow.
looking forward to seeing what yours looks like ugotd8. If your is different, then mine is unexplainable.
EDIT: the only way I could fathom it being a defect is if a large portion of the acetal wall is missing at threads. Hole cant be punched any more towards periphery. And if large portion is missing, and threads go all the way down, water is going to make it through a slit between bottom of threads and block.
EDIT2: I just posted on the alphacool forum support thread, so hopefully I will get an answer. It really looks like by design with sharp edges, and cant see any marred areas or how it could be miscut, but like others, my gut is telling me no one is that stupid to sell something like that on purpose...so have to wait and see what they say, or if no answer ...wait on pics from ugotd8 and confirm or not.
Interesting, with all the reviews on this block up, not 1 pic that has an angle where you could see the hole.
Yeah I guess all us reviewers were interested in a nice looking pic
Quote:
Originally Posted by
mdatmo
I think you forgot to include the sample size when calculating the confidence intervals. Confidence intervals are calculated from the standard error. The formula is
Code:
X +- 1.96*s / sqrt(n)
where n is the sample size, X the sample mean, and s the sample standard deviation.
The row should be
Code:
DT Sniper 41.40 0.25 0.25
Note: I am ignoring the fact that 5 is a very small sample size (statistically speaking) and thus should probably use the t-distribution, so we should preferably use 2.77 instead of 1.96.
I hated statistics so I'm glad someone payed attention in class
Quote:
Originally Posted by
opt33
For 1/2 hole, there are no threads at all, just about 1 mm of acetal. I did decide to test if would seal, after I finished using the block makers name in vain. I slowly screwed in the fitting (by the way it would be easy to cross thread I had to go very slow when starting to keep from doing so), it was just nearing finger tight, then became loose again, ie easily stripped. In 10 years of water cooling, countless builds and testing, never stripped a fitting or seen one leak, I always do finger tight then 1/8 turn with wrench. This effortlessly stripped with my fingers, and never anywhere near snug like the "in" hole fitting that I had put in first, per my pic above.
As far as them being on to something...no. The top design flaw or defect (dont know yet) has nothing to do with the fact they have one of the largest bases with a good broad bow and among most surface area of any block via 68 rows of cross cut pins 2.5mm high. The base is why it performs so well + jet plate design, not the fact it has 1/2 threads missing. Easy to fix that issue without altering performance, however might be few bucks more in cost.
Clearly whoever designed it knows the base is key along with jet plate. Because they were made decently. The rest of the block, like others mentioned, is simply very cheaply made obviously for cost cutting, phillips screws that will strip over time vs hex, mounting system is cheap crap, sticker plopped on top sometimes crooked vs engraved, etc.
No way would I even send for free (and obviously not sell) a defective design or defective part to another person. Although if it is just my top that is defective I will rma so I can test.
There is a third possibility, that it is both design flaw + defect, ie if my outlet hole was punched 1 mm in wrong spot plus acetal rim 1mm thinner, maybe threads go 3/4 way in others, though personally I still would not use even that. If that is case, and stren has a better top, if I decide not to rma since I wont use in my rig 24/7 anyway, might send to him so he can use his top with my base, so we can get another base vs base of same block for more variance testing.
For now, still waiting to hear from alphacool defect or design flaw. Though ugotd8 will let us know Tuesday, if his deliver comes as expected.
I'll be measuring all the bases and providing surface area estimates for each one once I take them apart.
Quote:
Originally Posted by
Lovidore
Really disappointed with the Summit EF performance. Specially considering I got one. -_- Under load at stock clocks reaches about 50 degrees C (about 22 degrees ambient? Give or take) Will consider getting a different block in the future.
Yeah I was too - I hoped it would do very well. Bitspower GPU blocks have similarly performed poorly in reviews though. Nice fittings though.
Quote:
Originally Posted by
opt33
Bitspower is one of the nicer blocks, aesthetic wise, and judging from pics and reviews, decent quality construction, but base has less surface area than many others and ? broad enough bow. It seems like when choosing a block there is ....quality construction, aesthetics, and performance. Manufactures can choose either one or two variables, apparently illegal to use all 3. That would also explain EKs choice to use circles.
I really like that block and was thinking of getting to test until read specs on base. Shame bitspower top/mounting doesnt fit the alphacool base.
It's got a huge base, much bigger than the acool. We'll have to see about the internals though.
Quote:
Originally Posted by
She loved E
just wanted to say thanks for doing this test. up to date info is always hard to find and thorough & up to date is even better!
selfishly speaking I wish the Raystorm copper performed better. Its hilarious that they charge more for it than the better performing std Raystorm. Well at least mine was hewn from pure solid metal.
Full metal blocks rarely improve anything siginificantly as they have minimal surface area to provide extra cooling. However the difference in material does affect the bowing of the block and this is usually behind any differences.
Quote:
Originally Posted by
mandrix
I'm in no way disputing anyone else's testing, but I have run both the copper and acetyl Raystorms on the same cpu/mobo and the results are close enough to call identical. Initially I thought the copper was better but after a few remounts of both they appear the same. I'm confident that for my testing on my rig the results are correct. All my testing was with MX-4 and MX-2 TIM.
Yes with only one test it's very hard to see any difference. Even with 5 tests of each block the results are close enough to possibly be manufacturing variation or test error.
Quote:
Originally Posted by
opt33
2 samples of same top raystorm have been shown to have a 1C difference measured with testing equipment. Other blocks have same issue, sample variation even using same top, ie 2 samples of exact same block. Strens testing of raystorm 1 testing 0.6C better than raystorm 2 is typical block to block variation, regardless of top.
Bow/base variation, milling of channel differences block to block are going to determine the variation in performance, so which top tests better will be primarily determined by which top got lucky and got the better base in that test.
So buy whichever top you like better.
@Ceadder, all Stren's testing is at load. If block 1 cools 1C better than block 2 at load TDP, then if idle TDP is 10x less, the difference is going to be closer to 0.1C, so idle measurements like you said, are worthless. Your probably questioning idle based on low numbers. But the temps shown are not core temps, they are core temps - water temps.
Yes idle is worthless, everything is tested at load.
Quote:
Originally Posted by
rrrhal
I stand corrected. Can you tell its been 30 years since I took those courses? There used to be a rule of thumb that 5 was the minimum sample size if you wanted to use this kind of analysis.
Ideally we would have at least 5 of each block and remount all 5 of them 5 times so we could run an ANOVA to see if there is more difference in how you mount the block, if you got a good example of a particular block, or if the design of a given block is actually superior. Then rerun these experiments at several constant flow rates. By now you can see we would need a small army of Strens to actually do this.
Well you're all better than me at this. I wish I could test more to get rid of a lot of the error, but like you say it's just too much time.
Quote:
Originally Posted by
Ceadderman 
Hope my edit is correct so that way when I say...
"Aha okay gotcha"... I can confidently know I came away with a better understanding an not look like a tard for making the adjustment.
Appreciate the clarification though.
~Ceadder
Lol
Quote:
Originally Posted by
cpachris
stren....does the mips iceforce block separate so that the "x" looking mounting bracket could be taken apart from the POM and be painted or powdercoated?
Glad someone answered you on this one chris. Yes the x bracket is totally separate and not held down in anyway. I actually wish it was held down, but it looks better this way.
I still owe you some photos on the white sniper.
Quote:
Originally Posted by
Blizlake
Employee maybe?
Lol maybe, maybe just trying to help. FWIW please respect the sponsors who gave the blocks to me so that we can have this data by not posting links to other stores! Soon I'll be posting a table with prices and vendor links to make it easier for people.
Quote:
Originally Posted by
Sixtron
Interesting test and thread. Reading here since some weeks so i sign up.
Thanks and welcome to OCN!
