I7 waterblocks Version2 (Ek Supreme LT | Koolance350 | Dtek Fuzionv2 | Swiftech GTz)

[Hondacity]

Introduction

The Intel corei7 cpu is a power hog of a cpu that needs watercooling. As everyone know already. I got the waterblocks purely for fun. And overclocking is my game. I wanted to try out which performs the best.

The test will be trying to cool the cpu as much as possible and doing so by watercooling. Using the dts sensors on the cores I will record them to get the best waterblock from the bunch.

Idle power of my mobo..and my full cpu load...this is a tremendous amount of wattage. I have the gtx285 when running it with furmark it only consumes 3xx watts from the wall. The core i7 cpu rapes it in terms of power usage. power meter: kill-a-watt

http://i24.photobucket.com/albums/c5/hondacity25/image003.jpg

http://i24.photobucket.com/albums/c5/hondacity25/image001-1.jpg



The setup

Radiator is the gtx480 from Hwlabs. Probably is the best 4x120 pc radiator available. The fans are Scythe Kaze 3000 fans in push pull configuration. Fans are powered from the Sunbeam rheobus extreme that has been overvolted to around 13v to achieve 12volts on the fans. Andreabz features the effect of push pull on the rads on his estimator (link). It is one of my favorite references. I would definitely call it the God book of Watercooling.

The pumps are 2 Swiftech 355 in series. Primary pump is with an Xspc reservoir top..then the second pump is with the regular xspc acryllic top. flow testing was done months ago. link to it:

http://www.xtremesystems.org/Forums/showthread.php?t=213106

1/2 true silver barbs and 1/2-3/4 tubes with some clamps.. distilled water silver coils.

The waterblocks

Koolance 350
Swiftech GTZ
DTek Fuzion V2
DTek Fuzion V2 with fits top (excluded - leaky)
EK Supreme (excluded – ran out of time)
EK Supreme LT

The simple game

Simple as it is. Here’s the brief procedure.

1. Connected everything
2. bled the loop and check for leaks
3. mounted the water block
4. load bios oc settings (4140 ht on vcore at 1.57v)
5. windows vista
6. realtemp 3…started logging
7. record room temp
8. prime95 torture test with the second choice large fft run 10-15 minutes, used large fft for more heat on the cores
9. repeat


Room temperature was controlled to about 23c, minimum and max temps were 22.7 and 23.1

The results

http://i24.photobucket.com/albums/c5/hondacity25/finaltemps.jpg

4th is the Dtek Fuzion V2. Similar to the GTZ, performance is ok, but it can’t play with the high restriction water blocks.

3rd is the Swiftech GTZ. Again it performed far from high restriction blocks. Here it shows how the core i7 overwhelmed it. when I tried this on the core2 series, it performed ok. take note of correct mounting.

Second place is the EK supreme LT, at first I mounted this block horizontal(the inlet and outlet were top to bottom) I was shocked how so so it performed, when i rotated it 90 degrees.To my surprise it performed really well. Flow seems really good just by watching how it bled. The delrin used is rather good. It bent alittle bit. flow testing will be done when i'm back.

-------------------------------------------------------------------------

First place is the Koolance 350, the first mount was there already since I tested it the last time...and did good. I used it primarily. Again showing its capability, it achieved the best temps. This remains my favorite.

:up:

[HESmelaugh]

Great stuff! I've been waiting for some numbers on the CPU350 and the Supreme LT. :)

Could you give some more details on the orientation of the Supreme LT? Because that's something I'm curious about.

http://www.abload.de/img/i7b0ki.jpg
Which would be the orientation that performed better, A or B?

[BringerOdeath]

So it looks like the LT is a good choice if you don't want to spend $35 dollars more for the k-350's slightly lower temps!

[Hondacity]

Great stuff! I've been waiting for some numbers on the CPU350 and the Supreme LT. :)

Could you give some more details on the orientation of the Supreme LT? Because that's something I'm curious about.

http://www.abload.de/img/i7b0ki.jpg
Which would be the orientation that performed better, A or B?

if you look at most x58 motherboards...the ek supreme lt performed well when the name on it is horizontal... i'll check later how the cpu is oriented based on your pics...

edit

http://i24.photobucket.com/albums/c5/hondacity25/P1020133-2.jpg

[N19h7m4r3]

Great stuff, I love my 350 even if its on an AMD 940 :D

[HESmelaugh]

Thanks for checking, Honda! And thanks for all the testing!

So, it's option "B" on my pic that performs better... that surprises me. I would have thought that flow along the length of the die would be better. Good thing you tested both options! :up:

[scamps]

omg - for what are all those fans? :shocked:

In which direction did you mount the 350 and the V2?

[Jhors2]

Just put one on order. Thanks :) Love the fans by the way, I hate turning on a computer and it feels limp wristed since you don't have the true blarring sound of a real machine :) hehe.

[Power_VANO]

Great job, Hondacity! I was waiting some numbers of CPU350 :) Thanks!

And as of flow testing we have already some info yes?

The least restrictive should be V2, than GTZ. compared to them, CPU350 is much more restrictive and I think LT should be somewhere between GTZ and CPU350.
So we got such results here yep? - more restrictive blocks have done better.

[SAE]

Wonder how the LT compares to the non-LT Supreme. Also wish you could compare the Koolance to Watercool's Heatkill Rev.3...
Great tests as usual, Hondacity! :up:

[Sadasius]

Wow! Did not expect that with the LT. It did better then the Fuzion V2. Well looks like I am buying another block....Damn you Hondacity.....making me buy another block.....:rofl: Thanks for the testing! :up:

[Mech0z]

Whats the LT standing for? Is this the same LINK (http://www.chilledpc.co.uk/shop/product_info.php?cPath=113_38&products_id=989) ?

[Sadasius]

Whats the LT standing for? Is this the same LINK (http://www.chilledpc.co.uk/shop/product_info.php?cPath=113_38&products_id=989) ?

It's the Supreme LT which is not the same as the original Supreme. Linky and pic.

Supreme LT (http://ekwaterblocks.com/shop/product_info.php?cPath=21_30&products_id=347&osCsid=625dcb1110f74d6ce45f526031a83548)

http://ekwaterblocks.com/shop/images/EK-Supreme-LT-Acetal.jpg

[SNiiPE_DoGG]

that supreme LT is a sharp looking block.... I may finally buy an EK cpu block :shocked:

[situman]

Nice job! Very informative. Looks like its time I upgrade from the V2. Does the orientation of the 350 matter as much as the LT?

[ballzD33P]

thanks for the time and effort, im sure alot of people will find this information valuable.
im interested to see how the normal ek-supreme does.

Also, have you ever used any of the cuplex blocks from aqua-computer? i'd really like to see how they fit into your testing, especilly the cuplex XT di
http://www.aqua-computer.de/e_index.htm

[Bojamijams]

Thanks for the test :)

I'm still impressed by the Fuzion v2.. sure its last, but its FAR less restrictive.. in fact from that, the GTZ faired the worst because it beat the fuzion only slightly but it'll rob a lot more of your flow

[Sadasius]

I am going to get this block and try it out once I get some numbers on flow first. If it's too restrictive I am not going to bother as I am cooling my whole mobo etc in the same loop. I am thinking a custom metal bracket for this block because I do not trust the delrin to hold when hard mounting with good enough pressure. I am sure the temps will be even better if you could put the screws to this without worrying about it bending and breaking.

[SNiiPE_DoGG]

thanks for the time and effort, im sure alot of people will find this information valuable.
im interested to see how the normal ek-supreme does.

Also, have you ever used any of the cuplex blocks from aqua-computer? i'd really like to see how they fit into your testing, especilly the cuplex XT di
http://www.aqua-computer.de/e_index.htm

martin has tested that block, and IIRC its not even in the same league as these blocks even on C2Q.

[Kashelz]

Thanks for testing, I was waiting to see what to get to raplace my old MP-05 SP LE block, now I have ordered Supreme LT, was thinking about Koolance 350, but it costs much more and no big advantage of it, so lets see how Supreme LT preforms in my configuration :up:

[shazza]

Nice job of testing ... thanks for doing this.

I'm okay to sacrifice a few degrees with my GTZ, due to the ease of mounting it - but I can see where the temp difference could be important to those pushing their systems to the max.

[sabe]

Ditto, I'm interested in a regular EK Supreme as well, so many people still swear by it :)

[DeanB]

Very nice... currently we're working on getting our restriction on the 350 down while maintaining temperatures. I've mailed two new impingement plates to Naekuh and Skinnie for testing :)

[N19h7m4r3]

Thanks alot Dean, sounds like great stuff. I already love my 350 but extra performance is always welcome.

[Stealth42o]

Thanks for the testing, I really appreciate it!

[ChaosAD]

Thanks for testing Handacity :up: I would like to go with the 350 but given the price of 80 euros vs 35 euros for the EK LT i think my vote goes to EK and it will be my next cpu block :)

[Hondacity]

omg - for what are all those fans? :shocked:

In which direction did you mount the 350 and the V2?

the fans? you're at xtremesystems :D

the v2 was tested horizontal..i doubt that rotating it will increase performance. but i will test that when i'm back also with the k350.


Nice job! Very informative. Looks like its time I upgrade from the V2. Does the orientation of the 350 matter as much as the LT?

probably not if you look at the internals..it looks like doesn't need special attention to it..but it will be tested


Thanks for the test :)

I'm still impressed by the Fuzion v2.. sure its last, but its FAR less restrictive.. in fact from that, the GTZ faired the worst because it beat the fuzion only slightly but it'll rob a lot more of your flow

i so agree here.. on core2 series my testing the dtek beats the gtz even martin had better temps with it..and better flow..


Ditto, I'm interested in a regular EK Supreme as well, so many people still swear by it :)

after several weeks...


Thanks for testing Handacity :up: I would like to go with the 350 but given the price of 80 euros vs 35 euros for the EK LT i think my vote goes to EK and it will be my next cpu block :)

thats why i wanted to test this quick..i know some would love to see results before their purchase...

[scamps]

Sorry, fans are for cooling down water and for nothing else :p: on xs and elsewhere ...

I am actually testing the 350 on a Q9550 @4GHz. And I can´t believe what I see at the moment: This cooler seems to be most restrictive one I ever had* :eek:
What are your results in this?

* incl. Supreme, Cuplex XT di, D-Tek Fuzion V2, Heatkiller 3.0 LT

[Blkout]

Good information to know, but since I already own a D-Tek Fusion, I just can't justify spending another $50-80 for a CPU block for 4c difference. Looks like I'll just buy a 1366 mounting bracket for the Fusion.

[NaeKuh]

increase heat load honda if your running dual pumps.

watch the heat load scale better on the KL-350. and i know your running dual pumps.

the block wont give you better temps at lower load, but it scales a lot better then the other blocks i tested with higher heat load.

this is why i kinda gave up after the third block testing i did personally.

[Hondacity]

higher heat load? i got a high heat load already... 250w approxiamately...

what was your heat load? if i increase the vcore higher...i get bsods

[NaeKuh]

higher heat load? i got a high heat load already... 250w approxiamately...

what was your heat load? if i increase the vcore higher...i get bsods

:rofl:

i was going as high as 1.55Vcore @ 4.25ghz and only noticed a 2C increase from 1.488

[msgclb]

Wouldn't it be better to judge by the maximum temperature excluding bad mounts. I've always found that my max temp is on Core0. I'd call it a tie between the Koolance 350 and the EK Supreme LT.

95496

[scamps]

I finished my test on a Q9550. Perhaps you should add the Watercool Heatkiller 3.0 to your review. If it is as much better on a i7 as it is better on a Q9550 ....

My little review could be find here (german language only):
http://www.forumdeluxx.de/forum/showthread.php?p=11454795#post11454795

http://www.abload.de/thumb/auswertungmediumefmb.jpg (http://www.abload.de/image.php?img=auswertungmediumefmb.jpg)

[Hondacity]

Wouldn't it be better to judge by the maximum temperature excluding bad mounts. I've always found that my max temp is on Core0. I'd call it a tie between the Koolance 350 and the EK Supreme LT.



it looks almost the same..if i remount the k350 several more times..i'd probably achieve lower temps...thats why i took the lowest temps..because its achievable..if you consider the max temps then realistically you're disregarding the better capability...

[mstrmold]

Great writeup Hondacity. Question for you on the GTZ, how was it mounted? Seems there is some debate on whether or not you should follow Swiftech's guidelines with lining the swiftech logo up with the release latch of the 1366 socket. I'm getting ready for my i7 build and wonder which was the best mount orientation for you?

[Envy]

high voltage i7:"This is MADNESS!" :shock:
high overclock".........THIS IS SUPREME LT!" :upset:

http://www.saunders.fr/images/oc22.jpg
http://www.saunders.fr/images/02.jpg

[warriorpoet]

Your fits top leaked!?!?!?!

[CedricFP]

So does the Supreme LT perform better than the Supreme? IF so, why is it cheaper?


Further, what are the mounting mechanisms of the KL 350 and LT? Is it a hard mount?

[Hondacity]

Great writeup Hondacity. Question for you on the GTZ, how was it mounted? Seems there is some debate on whether or not you should follow Swiftech's guidelines with lining the swiftech logo up with the release latch of the 1366 socket. I'm getting ready for my i7 build and wonder which was the best mount orientation for you?

i followed the instructions..which gave better temps than not following instructions...


Your fits top leaked!?!?!?!

yes


So does the Supreme LT perform better than the Supreme? IF so, why is it cheaper?



Further, what are the mounting mechanisms of the KL 350 and LT? Is it a hard mount?

i'm not sure yet..but design wise its definitely cheaper...remember the xspc gpu full cover water blocks...

performance wise..i'd guess the LT will perform better...i have yet to mount the ek supreme

i used hard mount i get better mounting repeatability with my hard mounts:D

[Hondacity]

high voltage i7:"This is MADNESS!" :shock:
high overclock".........THIS IS SUPREME LT!" :upset:

http://www.saunders.fr/images/oc22.jpg
http://www.saunders.fr/images/02.jpg

what is that case? the rad mounting is beautiful...where is the harddrive? dang i need to beat your score...i only got 8.987:up:

[NaeKuh]

high voltage i7:"This is MADNESS!" :shock:
high overclock".........THIS IS SUPREME LT!" :upset:


:rofl:

This is the KL-350 done right owning on Linpack:

http://i125.photobucket.com/albums/p73/aigomorla/Final-1.jpg

:rofl:

HT on Also.

Im gonna say this, for us xtreme people, i have yet to see a block out perform the KL-350. But you need the setup tho.

Wanna see a 920?
http://i125.photobucket.com/albums/p73/aigomorla/Linpack.jpg

[SNiiPE_DoGG]

naekuh your only at 4.1ghz with a much lower voltage....

[Vapor]

Two different CPUs, can't really compare anyway :shrug:

[Crazystang01]

Thanks for doing this write up I purchased a EK Supreme LT on blind faith before this review. Hopping it would preform

[NeedMoMegaHurtZ]

Thx for sharing the testing results! Very interesting...

[Petra]

it looks almost the same..if i remount the k350 several more times..i'd probably achieve lower temps...thats why i took the lowest temps..because its achievable..if you consider the max temps then realistically you're disregarding the better capability...

Just because it's achievable doesn't mean that it's representative of the product in the larger scope and simply focusing on the best that you can achieve while throwing the normative data out does not make for good testing methodology.

Even with your data averaged, you're still suggesting that the difference between the 350 and the v2 is akin to the difference between one of Stew's Storm G5's and an MCW5000 (to pull a comparison from the past for the purpose of focusing on magnitude of design difference). Your spread doesn't seem at all odd to you, given the typical spread that controlled testing yields?

You have some reading to do and, since you don't seem to like hearing it from me, here's a good place to start (old but still plenty educational):

http://www.procooling.com/index.php?func=articles&disp=138
http://www.procooling.com/index.php?func=articles&disp=140&pg=1

[NaeKuh]

naekuh your only at 4.1ghz with a much lower voltage....

HT ON + Linpack vs superPI?


Two different CPUs, can't really compare anyway :shrug:

yeah i know,

:P

But i wanted to just throw it out.

[DeanB]

Wow...Just got my 350 in today. I'm impressed to say the least on parts quailty and construction. I know one thing, you'll never have to worry about the hold down or backing plate flexing. That's for damn sure.:D I'm sure it's some where, but what's the proper orientation for a i7 920?

Can't wait to get this bad boy mounted. What a difference a year can make with a LC company, very well done.:up:

andyc

Thank you :)

You can skim the optimization guide (http://www.koolance.com/support/files/koolance_cpu-350_optimization.pdf). Parts 3 and 4 mention correct orientation.

[shazza]

You guys sure have peaked my interest in the 350. Especially since my CPU is running in it's own loop with a PA 120.3. And ... Naekuh, that 965 looks fantastic. Is that just on water, or are you using something else as well? (sorry to derail the thread a bit, but this is good stuff).

While we could pick Hondacity's testing apart (e.g. run an analysis of variance on the individual data points, and you may find the statistical significance of the differences is less than you might think based on averages or the lowest number achieved) ... I really appreciate someone doing this. I cry every time I have to reseat my CPU block, so I appreciate the effort that goes into this.

[Nemon]

Very nice review, Thanks!

I installed the Supreme LT last night and the temps are the same as the KL350. idle and under load.

Its a good spare block for me now, will be switching back to the KL350, I like the clear top:D


http://i28.photobucket.com/albums/c239/vkvenom/EKLT.jpg

[BringerOdeath]

Very nice review, Thanks!

I installed the Supreme LT last night and the temps are the same as the KL350. idle and under load.

Its a good spare block for me now, will be switching back to the KL350, I like the clear top:D


http://i28.photobucket.com/albums/c239/vkvenom/EKLT.jpg

You can buy the acryllic top one from EK! Although I am not sure how strong the acryllic would be for mounting.....;)

[Hondacity]

OK, you guys are going to love this one.

I'm sure it's something simple that I've overlooked, but what attaches the backing plate to the MB prior to mounting the block? My bench is horizontal with a cutout on the tray, so if nothing holds the backing plate to the MB, it falls out all the way down to the PSU:rofl:

andyc

:rofl:

i use my own screws:up: i clamp it to the motherboard so remounting is a breeze..:yepp: have fun:up: can't wait to see your numbers

[scamps]

@Hondacity
Please post some information about your flow, I would really like to compare them with mine.

[Hondacity]

my flow testing

http://www.xtremesystems.org/Forums/showthread.php?t=213106

[FragMagnet]

You can buy the acryllic top one from EK! Although I am not sure how strong the acryllic would be for mounting.....;)

Not very well, I had a GeminiCool on a P4 years ago, it cracked, so did the replacement, and the fabricated ones thereafter. Can't say for sure if it was acrylic, but it was a clear plastic. It cracks near the fittings, even with plastic fittings. I will never touch a WB with plastic used as a structural part again !

But it did give me justification for the wife to buy my Mach 1 !

[scamps]

my flow testing

http://www.xtremesystems.org/Forums/showthread.php?t=213106

Thanks a lot, I haven´t noticed yet. So my (in my opinion: bad) results seem to be right :eek:
I even wondered, if I could have get a dickey one ...

[NaeKuh]

I'm really starting to like this block,

andyc

Jumped kinda late there bro, but your finally here.

The block scales very nicely with heat load as i was saying. Your gonna love the block, and it loves pressure and flow.

I got a few injectors which im gona send out to bei to get some custom drilling done on. If i like a config, i'll make sure i pass one to ya since dean gave me a few.

:up:

[santiagodraco]

Jumped kinda late there bro, but your finally here.

The block scales very nicely with heat load as i was saying. Your gonna love the block, and it loves pressure and flow.

I got a few injectors which im gona send out to bei to get some custom drilling done on. If i like a config, i'll make sure i pass one to ya since dean gave me a few.

:up:

I think this is what impresses me most about this block. I've been running the 350 in a single loop with my 4870x2 GPU block with great results under full CPU and GPU load. I'm planning on moving to i7 soon and adding an additional GPU to the loop, but might wait for the next gen ATI cards before I do that. We'll see!

[rioja]

honda, how did u mount d-tek? i saw somewhere u used a 1366 mount from ice-hammer, right?

[Sadasius]

honda, how did u mount d-tek?

Well first he bought it dinner.....:rofl:

[rioja]

i have to decide should i change my fuzion v1 with v2, to make smth like this
http://i155.photobucket.com/albums/s314/Salapao2000/DSC00472.jpg

or leave v1 and use ice-hammer round mountings
(can't find those pic)
any advice?

sorry for o/t

edit: i have found, d-tek v2 with ice hammer bracket
http://www.xtremesystems.org/forums/showpost.php?p=3639753&postcount=9

[DarthBeavis]

:rofl:

This is the KL-350 done right owning on Linpack:

http://i125.photobucket.com/albums/p73/aigomorla/Final-1.jpg

:rofl:

HT on Also.

Im gonna say this, for us xtreme people, i have yet to see a block out perform the KL-350. But you need the setup tho.

Wanna see a 920?
http://i125.photobucket.com/albums/p73/aigomorla/Linpack.jpg

About to start doing some Ocing on this puppy. I heard you need less Vcore for more love . . .we will see. Conundrum should also have one by now (I have two of them).
http://img9.imageshack.us/img9/4416/cpuzt.jpg

[Jhors2]

Hey honda, what were the temp differences between running one pump and running two?? just out of curiosity. Thanks!

[Hondacity]

my memory has bad blocks...sorry i don't remember..but more flow should help lowering the temps for the high restriction blocks...