PDA

View Full Version : Stepped Cuplex Anyone?


NaeKuh
10-13-2008, 11:30 PM
Hehe

Teasers, got my block back from martin.

<3!!! :up:

i'll think about posting the results... :rofl:
Well martin hasnt given them to me yet also. :rofl:

Provided my friends also dont slaughter me soon for holding them...

Anyhow the teasers:

http://i125.photobucket.com/albums/p73/aigomorla/IMG_0061.jpg

http://i125.photobucket.com/albums/p73/aigomorla/IMG_0062.jpg

http://i125.photobucket.com/albums/p73/aigomorla/IMG_0063.jpg

<3 customs.
twwen2
10-13-2008, 11:40 PM
Very nice my friend. :yepp:

Now get those results off Martin or imma reach for my torch and pitchfork! :p:
septim
10-14-2008, 04:42 AM
yeah neakuh, please share some results from those test you and martin had. it would be an eye opener to block making companies...
WaterFlex
10-14-2008, 05:19 AM
Amazing! Like it very much. Some results, please.
Broken
10-14-2008, 05:48 AM
Yes, I wantz resultz!
Eddie3dfx
10-14-2008, 06:55 AM
Looks gorgeous.
Can't wait to see what kind of temps you get :up:
71 (Bryan)
10-14-2008, 07:54 AM
srry for not neing online for a long time but this is something i wanna see
nikhsub1
10-14-2008, 08:02 AM
yeah neakuh, please share some results from those test you and martin had. it would be an eye opener to block making companies...
And why is that? I have been saying that the step is beneficial for over a year and a half. I understand your point though, I'm mostly ignored.
aajj7733
10-14-2008, 08:35 AM
Hi
What are the measurements of the stepping?
mine is heading towards a CNC machine towards the end of the week anyway so was wondering if this is worth doing?
thx
Eddie3dfx
10-14-2008, 09:03 AM
Hi
What are the measurements of the stepping?
mine is heading towards a CNC machine towards the end of the week anyway so was whomering if this is worth doing?
thx

Yes, of course.
skinnee
10-14-2008, 09:07 AM
NaeKuh...put that thing in your safe. If gillbot doesn't get it, I might just have to take a "recon" trip to Cali. :D
T_Flight
10-14-2008, 09:27 AM
martin, if you read this, very nice wiork with the parallells, and milling. It looks very true.

nikhsub1, trust me, you're not being ignored. I have read lots of great info from you here at this forum, and it has helped me alot. Thanks. :)
Boogerlad
10-14-2008, 12:30 PM
it is 23*23mm
NaeKuh
10-14-2008, 12:52 PM
no i had mine match the internal pin matrix size.

so its 25x25.

hmmm... results... results... whats that? :rofl:

I'll give martin the OK to release it. possibly... :rofl:
BlueAqua
10-14-2008, 01:19 PM
NICE!! I want one! ;)

I wish I could step my incoming Cuplex DI HD.
aajj7733
10-14-2008, 02:03 PM
what is the height also please
0.5mm? 1mm?
thx
Martinm210
10-14-2008, 02:42 PM
I thought I sent them to you, sorry. Here are the results:

Before:
http://img55.imageshack.us/img55/776/acdistockgf3.png

After:
http://img55.imageshack.us/img55/5441/acdisteppedwg9.png

Compared:
http://img55.imageshack.us/img55/3425/acdistockvssteppedlr1.png

On average it was worth about a .5C gain primarily with the average of all four cores...that was worth it I think..:up:
twwen2
10-14-2008, 03:16 PM
And why is that? I have been saying that the step is beneficial for over a year and a half. I understand your point though, I'm mostly ignored.

Don't take it too hard scott...:(
NaeKuh
10-14-2008, 03:43 PM
ack btw martin, your wolfie went out today. :rofl:

You should have it to play around with soon. I'll be welcoming you to the 4ghz club once u get it. :rofl: :up:
gillbot
10-14-2008, 04:37 PM
NaeKuh...put that thing in your safe. If gillbot doesn't get it, I might just have to take a "recon" trip to Cali. :D

Too late, I already got it. ;)
philbrown23
10-14-2008, 05:13 PM
man not to do this with my apogee GTZ :P
NaeKuh
10-14-2008, 05:20 PM
man not to do this with my apogee GTZ :P

you cant, unless you lapped the bowed base down flat b4 you stepped it.
Bojamijams
10-14-2008, 07:15 PM
Am I the only one that finds these results extremely unimpressive and easily attributed to user error? Hardly seemed worth it :(
Martinm210
10-14-2008, 08:19 PM
ack btw martin, your wolfie went out today. :rofl:

You should have it to play around with soon. I'll be welcoming you to the 4ghz club once u get it. :rofl: :up:

Awesome, should be fun.
Just in time for my new tech station build...soo many toys...:D
http://img529.imageshack.us/img529/2740/torturerackyl5.jpg
Helfarch
10-14-2008, 08:37 PM
Am I the only one that finds these results extremely unimpressive and easily attributed to user error? Hardly seemed worth it :(

Graphs are nice and all but it's not even a whole degree. Unless you happen to have your own CNC machine it's not worth it and the extra machining costs probably aren't worth it to a manufacturer either.
skinnee
10-14-2008, 09:28 PM
Awesome, should be fun.
Just in time for my new tech station build...soo many toys...:D
http://img529.imageshack.us/img529/2740/torturerackyl5.jpg

You're going to love the setup on a torture rack! I am using the BlueAqua Custom Series :up:, and its perfect for running tests...especially block mounts.
NaeKuh
10-14-2008, 10:17 PM
skinee, i should be getting my banchetto VERY Soon. :rofl:

makes the torture rack look cheap. :rofl:
MomijiTMO
10-14-2008, 10:32 PM
Well the torture rack is cheap compared to the banchetto :D
twwen2
10-14-2008, 10:34 PM
skinee, i should be getting my banchetto VERY Soon. :rofl:

makes the torture rack look cheap. :rofl:

Doesn't it ever!
Those benchetto things look divine. :D
skinnee
10-14-2008, 10:36 PM
skinee, i should be getting my banchetto VERY Soon. :rofl:

makes the torture rack look cheap. :rofl:

You guys and your acrylic...wood is where its at! Want to mount a pump somwhere...break out the drill! :rofl:
NaeKuh
10-14-2008, 10:49 PM
You guys and your acrylic...wood is where its at! Want to mount a pump somwhere...break out the drill! :rofl:
Only someone who hasnt seen the beauty of a banchetto would say something like this :rofl:

http://www.vr-zone.com/articles/Microcool_Banchetto_101_Case_For_Enthusiasts/5980-2.html

http://resources.vr-zone.com/newvr/image.php?m=600&s=http://resources.vr-zone.com//newspics/Aug08/12/3d_retro.jpg

im gonna mount my 480 up front here:
http://resources.vr-zone.com/newvr/image.php?m=600&s=http://resources.vr-zone.com//newspics/Aug08/12/Banchetto/3d_retro_unit_centr.jpg

and next to it will be a 3/8 recip. H2o Chiller :P


Thats gonna be eva-chan.
Bwahahahahahahahahaha... :up:
skinnee
10-14-2008, 11:20 PM
Oh I've seen it...just acrylic and DICE benching are not a good mix.

It is a great looking setup, just not what I ended up really wanting. Plus, I got a custom now, a BlueAqua custom! :D
BlueAqua
10-15-2008, 07:27 AM
I'm partial to the BlueAqua custom myself. I still haven't seen a test bench as tough and functional as that one. I'm glad you like it Skinnee, it cost me some serious change.

BTW, WTB Stepped Cuplex. ;)
WaterFlex
10-15-2008, 10:00 AM
NaeKuh
Wow! Nice staff...
kinghong1970
10-15-2008, 10:05 AM
@Naekuh:

dude, where do you get one of those?....
/fiddles with credit card...

@WaterFlex...
um.. what are you doing lookin at NaeKuh's "staff"?
Waterlogged
10-15-2008, 10:06 AM
I think there are still gains to be found in NaeKuh's block. He went with 25mm˛ when it should have been 23mm˛. That extra 2mm could allow better pressure on the cores and we could see an extra full degree or more at that point on top of what he got.


kinghong, Try Sharka (http://www.sharkacomputers.com/aqcodoimin.html) ;)
nikhsub1
10-15-2008, 12:22 PM
I think there are still gains to be found in NaeKuh's block. He went with 25mm˛ when it should have been 23mm˛. That extra 2mm could allow better pressure on the cores and we could see an extra full degree or more at that point on top of what he got.
Im glad you said it first... I balked at the 25mm size step because it is too big, it rests on the supports on the underside of the IHS... really no better than not stepped. Yes I do think 2mm makes a big difference in this respect.
NaeKuh
10-15-2008, 12:46 PM
Im glad you said it first... I balked at the 25mm size step because it is too big, it rests on the supports on the underside of the IHS... really no better than not stepped. Yes I do think 2mm makes a big difference in this respect.

yeah but i compensated for the larger IHS coverage on i7.

So that is why i told martin to keep and hold the pin matrix.


Successful or failed, it was fun for martin and me, so i got what i wanted out of it. :rofl:
Martinm210
10-15-2008, 03:23 PM
Yeah, it was an interesting test.

Worth is all relative. It didn't cost me or Naekuh anything but shipping and a little time, so cost was relatively nothing. No arguing against the fact that .5C or even 4C is going to make any sort of difference in overclock potential, but regarding design and pushing the envelope on performance...why not. All the blocks out there are within about 3-4 degrees tops, and the AC DI is a strong performer.

This .5C step managed to bring it from something behind the GTZ to something that's in the same ballpark.

Personally I still think bowing is overall better than stepping, but it depends on the block itself. The more ridgid blocks like the AC DI having a metal top probably perform better with stepping than some of the plastic top blocks that flex. If the block is bowed and the block flexes, there's still little chance you'll ever start bridging the corners on the IHS. But if the block is stepped and the block flexes, you now may risk bridging the block over the corners of the step.

I think what is really comes down to is pressure per unit area. Stepping is going to increase that because of the smaller area, but it's still a very distributed and even loading. Bowing on the other hand will progressively increase the pressure toward the center of the IHS and reduce slightly as you move away from that point.

For round numbers, look at it this way. Assuming the block is ridgid.

IHS = 1.16"(about 30mm) x 1.16" = 1.345 square inches. If you mount with 50 lbs of force, your force per unit area is 37.16PSI

25mm step = .983" x .983" = .9662 square inches. 50 lbs of force = 51.8PSI which is about a 40% increase in pressure over stock

23mm step = .905" x .905" = .8190 square inches. 50 lbs of force = 61.1PSI which is about an 18% increase over the 25mm step.

The problem is you are also loosing contact of the IHS at the perimeter and while it isn't nearly as important as the center die area, it does dissipate some heat on quad cores. I didn't really notice it on dual cores, but after testing all these blocks over 5 mounts, I became religious about making sure I had full TIM spread over the entire IHS or I would generally see worse temperatures on the Q6600. The E6600 didn't seem to care about covering the IHS nearly as much as the quad, so I think there is some differences depending on the processor die.

This is also why I think bowing is better, even though you focus most of your cooling at the die locations, you still have contact with the perimeter of the IHS to pull away the little bit of heat there is there.

Anyhow, I could be wrong, but I think bowing is better than stepping, it's just that some blocks are pretty hard to bow by nature of their design. That's where stepping comes in...
CyberDruid
10-15-2008, 03:51 PM
Interesting.

Welcome to the stepped CPU club.

I've got two of them from Iandh :D 23 x 23 of course. Seems to make a slight difference...and I guess that's what we waterheads look for...slight differences.
cegras
10-15-2008, 05:13 PM
For CPUs with soldered IHS though, how can it make that much of a difference? I'm sure solder is somewhat soft, but to even deform the IHS to a point that you reduce the die-solder-IHS interface?
nikhsub1
10-15-2008, 06:27 PM
For CPUs with soldered IHS though, how can it make that much of a difference? I'm sure solder is somewhat soft, but to even deform the IHS to a point that you reduce the die-solder-IHS interface?
http://i86.photobucket.com/albums/k93/chrishorgen/Funnay/facepalm.jpg
nikhsub1
10-16-2008, 07:59 AM
If you are unsure what a bow or a step does, please read these.

http://www.xtremesystems.org/forums/showthread.php?t=142998 (Bowing)
http://www.xtremesystems.org/forums/showthread.php?t=160566 (stepping)


This pic ought to light the bulb up. (Pic Courtesy of David from cooling-masters.com)

http://www.cooling-masters.com/images/articles/apogeegt/images/contact.png
NaeKuh
10-16-2008, 08:13 AM
If you are unsure what a bow or a step does, please read these.

http://www.xtremesystems.org/forums/showthread.php?t=142998 (Bowing)
http://www.xtremesystems.org/forums/showthread.php?t=160566 (stepping)


This pic ought to light the bulb up. (Pic Courtesy of David from cooling-masters.com)

http://www.cooling-masters.com/images/articles/apogeegt/images/contact.png

scott we should make a new picture of that with this:

http://i125.photobucket.com/albums/p73/aigomorla/IMG_0063.jpg
Kayin
10-16-2008, 08:39 AM
Naekuh, since nobody else has done it, indulge me for this one.

OMG, that block has acrylic in it, we're all gonna die!

I apologize, but it had to make it in here somehow, and at least it could be me saying it tongue in cheek and knowing full well you know how to treat it instead a n00b.

Interesting on the 25x25 choice, but I think with Bloomfield it will pay off. Larger IHS will mean that will fit perfectly. Might be the go-to block for Nehalem.

Now to get pics of my (homemade) DI up. When it finally seals.
cegras
10-16-2008, 10:05 AM
4chan motivation poster lulz

Alright, since apparently I can stand to learn from this, I have visited each thread and skimmed the entire thing.


Think of the dynamics of an IHS for a second. Have you ever seen the underside of it? Well it has a ridge around it for support round the outer edge. If you can apply pressure INSIDE of that ridge you can get more pressure on the die(s).

What. Unless the laws of physics are broken, where is your pressure going on a stepped block? Upwards? Diagonally? Sideways? Or maybe, downwards, into the IHS over the die, deforming it and pressing it closer to the die? Unless your cpu package is made out of some physics-defying material that doesn't deform and all you're trying to do is lessen the TIM interfacial. Perhaps surface mating has some effect, but it's pretty clear that even the sum of all these three factors leads to a 0.5C measured difference.

http://i35.tinypic.com/2m6r2i8.gif

Where is the pressure going?

Well, I see you have an attitude to everyone on the forum, so I'll brush off your 4chan /b/tardism as it stands.

(Here I recognize that most 'veterans' will start to rain abuse on me, so I'll start by apologizing in advance if the thread derails. Understand that 'famous' or not, I'm sure everyone can get hurt feelings .. especially people with huge egos.)


Those of you who have known me for a while know that I am insane - I believe I single handedly started the 'Overkill' revolution.

Funny, I find the people that proclaim they are weird and / or insane are the ones that are the most boring. The only way to be truly weird or insane is to behave in a socially deviant way and find it totally normal yourself.