The IHS temperature is NOT important to overclocking, which is what Cathar is saying.

I believe that the IHS temperature is used in testing because it's an easily acquired number - you can't drill a hole in a CPU, but you can in a metal plate above it. IHS based testing will not indicate what the CPU temperature below is, but that's not the goal. The goal is to see whether Block A will produce a warmer or cooler CPU than Block B. Measuring above the IHS also removes the IHS-CPU TIM joint variable, so you don't have to worry about that degrading.

But it is usable as a method for establishing heat exchanger performance.

However, as with any method the relationship to normal reality needs to be understood.
I'm trying to model this kind of behaviour.
There are differences between blocks on the TTV (and all test testups) that don't exist in reality. This is true with or without the IHS
Some modelling which I'm too fatigued to explain at this point.
Not visible in the gradient chart but there is a 0.02° difference between the two blocks in the two different scenarios at the die core, meaning that the test setup is itself dependent on the block it is testing.
This is exacerbated by TIM variations etc. which I have not modelled at this point. (too complex in a hurry) and even more significant with thin vs thick bases, concentrated cooling regions bla bla

Any test method is inherently flawed unless we understand it.


I am almost typing in my sleep here, forgive the incoherence.

Does anyone know if it was testing with 3/8" barbs or 1/2" barbs? I'm guessing 3/8" because it was compared to the 6000, but I'm still wondering.

http://www.swiftnets.com/assets/imag...%20vs%20FR.GIF

says it is compared with mcw-6002

the test data in question has nothing to do with the overclockability of a cpu. it demonstrates the components' thermal resistance. thermal resistance is our standard of measuring thermal component performance, no? the goal of testers is to have accurate results for c/w, right? i know cathar takes great time in explaining a lot of things to the rest of us, but i still can't understand that because the system is not measuring die temp it results in less accurate results for c/w.

Did anyone read my second paragraph? :(

hah, no, sorry, my apologies.

I've been out of the "scene" for a bit, but unless things have changed, aren't enthusiasts expected to remove the IHS?? Are we really such babies now that we can't deal with bare dies? (Ironically, I chipped a die in my last watercooling setup, but I blame it on a poor mounting system ;) ) Or are waterblocks now attempting to market themselves as for the average "plug'n'play" user?

Unless the goal is a truly mass market block, then all testing should be done on a bare die.

My assumption is that most "enthusiasts" are still going for every extra C° and will not hesitate to remove the IHS. But then it really depends on how you define the enthusiast market and how you characterize the average enthusiast.

IMO, this is PROCOOLING, which implies going somewhat further beyond the average "amateur" attempts at cooling. Other communities may have different standards.

Obviously we are now at the point in our CPU water cooling history where the market may be split, where certain blocks are designed around the IHS, and other designs, catering to the "enthusiast" market, will assume a bare die. So in that sense, there will be two seperate criteria for "best" block. Designers should state clearly which they are aiming at.

on intel chips, the IHS is epoxied on, no way to remove it without destroying the cpu
or at least no one has found a way
but then i guess you can argue "true enthusiasts" only buy amd
i think the IHS is becoming a reality that even the enthusiasts can't avoid. we will have to deal.

Guys, I think that perhaps some of what I'm saying is getting lost in the jargon.

My point is that I don't believe that the TTV is providing a uniform heat flux when the block geometries change (different waterblocks) and this correspondingly affects the measured/reported IHS surface temperature. I am saying that unless we also know the die temp, then we just really don't know what the deal is.

If the heat-flux is guaranteed to be uniform, such as the heat-die and the IHS being an integrated solid one-metal-piece affair, then it could be used as it presently is, and the assumptions would be valid that the IHS surface temp and the die temp is a linear invarying relationship.

sorry but this is blah blah stuff
Cathar, you are now consistently stating that a die temp is necessary to evaluate an 'enthauast' wb
- as jd would say; put up, or shut up

again, please describe the acceptable heat source
no more bs, tell us what is required

is such presently available ?
what is its accuracy and repeatability ? (in terms of watts and temps)
define your 'heat flux' uniformity requirements, and the measurement method and equipment

from your post above, I am at a loss to understand who meets your requirements
other than that no one does it correctly
talk real world, not what if

you are attempting to pull the WCing community 3 steps backwards by saying only an actual test on an (unquantified) CPU will reveal the mysterious inner capabilities

this is crap, data needs a correlation to be useful - that is what you should be questioning

Sorry Bill,

The moment you can come up with a suitable explanation for the TTV data for the stupendously crippled XP, a block which has an even smaller main cooling patch than a Storm, then I'll for 1 moment consider that the TTV's die->IHS TIM layer is invariable.

You've seen the data. I at least attempt to make an explanation for the reasons for it, rather than sticking one's head in the sand and blindly sticking to one's guns in the face that what is seen is simply and utterly not possible, both in theory and in practise.

This has come down to one simple basic point. I refuse to accept that which is theoretically impossible. You can't change the laws of physics. You can't just blindly accept that a well known and analysed design (re: the XP) is achieving an h(eff) that is now projecting to be between 5-10x higher than that which was ever thought possible, or has an h(eff) 5-10x higher than anyone has ever independently shown, and the ONLY new data point is the TTV that suggests this.

As it stands, the Apogee's results with the 3mm base-plate would need to have an h(eff) that is around 2x-3x that which has ever been independently speculated for any design, anywhere, for the empirical TTV results to be true.

I utterly, completely and catergorically deny that some waterblock design that is at heart minimally different to those which have come before it (re: MCW5000), has suddenly extracted a ~10x h(eff) improvement with what is really quite minimal design changes.

Can change a lot of things, but can't change the basic laws of physics. H(eff) between the small die and large die results shown to date all form an incredible disjunction the moment the TTV data is put into the picture. In prior history, all results roughly correlated.

Why?

I AM asking the questions. Problem that I see here is someone who very stubbornly doesn't like to even suggest an answer, who is postulating that nearly 100 years of theory should suddenly be thrown out the door because Intel made a TTV. Sorry, I don't buy it.

So, why not just use a copper slug shaped like an athlon 64 x2 die set up to give off 100 watts or whatever makes you people tick? Seems a lot simpler and its one of the largest dies we'll encounter for a long time (possibly ever)

All that matters is die temp, regardless of what is between the waterblock and the die. Our final goal is to cool the die; to measure the temp of something else to determine whether or not a cooling solution is any good is purely insane. The purpose of our cooling solutions is to get die temperature as low as possible, not IHS temperature. True, IHSes are becoming a fairly permanent item, but they still don't matter in a block's performance. Also, even Intel says that the TTV is for one thing: determining whether or not a heatsink will keep a chip from overheating or not. Not determining waterblock c/ws to within.1 degrees C. Even Intel says that other systems should be used in addition to the TTV to simply certify a heatsink as not-gonna-nuke-your-system.
Bill, why are you so hopelessly defending a mediocre system that doesn't even give us what we want? All it does is cast doubt across all of Swiftech's numbers, doubt that is clearly exemplified in a 6-page thread about a new block, 5 pages of which are a debate about whether or not the numbers are meaningful.

Oh.
See attachment


MCW6000 60x60x5mm bp 281 (9x2x2mm) pins with 6mm dia nozzle ?
Modelled as 60x60x5mm bp, 289(9x2x2)pins in17x17 array, 6mm diam nozzle?
Any one?

Edit1: Added Apogee 10mm ID

Edit2: Added graph for predicted performance on 1sq cm Heat-Die
Edit3: Added Predicted "h(eff)" Profiles
Edit4: Added Predicted Performance on 32x32mm Heat-Die

I do share your concerns with the validity of the data (as it is not mine), I assume the equip is being used correctly - not unrealistic based on my exp

again, you have not proposed a solution to what you are calling a problem

and yet again Cathar, I am an experimentalist (Thanks deepow) and try to generate good data
that you cannot fit some of the data into your analysis suggests 2 possibilities:
- that the data is not intelligble / understood (limits of applicability, etc.), or
- that your analysis is misusing a datum not so intended or derived

you are pounding on me about your analysis when I respond to data
I am asking for heat source suggestions which you ignore,
your revenge for my ignoring your analysis I suppose

BS
what I am "hopelessly defending" is not Intel's TTV, it is the ability to extract data from a test bench
-> I do not agree that exact simultude is necessary

the worst case SIN of which Swiftech can be guilty is perhaps not understanding a needed correlation,
another variable got added and comprehension went to zero

you guys are making a VERY good case for mfgrs to NOT PROVIDE DATA
Swiftech is to be blamed for "causing confusion" ? (did they not describe the testing ?)

I guess Cathar and his followers are jockeying for position to be able to say I told you so, just in case

loverly, I suppose
is this an analysis contravening the laws of physics ?

1)Colours could be better but stuck with original data colour scheme
2) Not the laws I use.
It is the dimensions of Resistivity which are critical.
Dunno the raw data from TTV
But could easily be represented as relative to Vehicle size or as I have as C/w/cm^2

clearly a size 'factor' is needed, (the readers may imagine if this was overlooked by the heat source mfgr)
I prefer °C/W/cm² for the units as it forces addressing the issue

cm^2 are you japanese? those are some odd units :P

Most of the stuff i said in this thread should be ignored, as i have made some bad assumptions (base thickness etc). I still dont believe the apogee can beat the storm style but we will see.

The stuff on CFD is wrong as i thought that the block used jet impingement. Probably they got COSMOS out and parameterised the design to optimise it, im not sure how accurate its going to be on such small scales as you need alot of elements to get things right. Based on the design of it this could well be as good as your going to get with pins. They've clearly thought abit about it, as a DFMA (design for manufacture and assembly) piece it looks really good and cheap.

I always wondered why stew didnt injection mould his top pieces there might be issues with accuracy but could quite easily lower prices.

There are discussions on OCAU about such techniques but the volumes required to offset the setup costs where not seen viable.

It's believable in the same way that someone who tells you that they've just constructed a car that looks just like every other race car, has roughly the same characteristics as any other race car, but it claimed that it can lap some racetrack in 1/5th the time of the best lap record present.

Potentially possible: yes. Believable: no.

but excellent mktg, look at the buzz
and more to come for sure

Can only cry wolf a few times before people catch on.

As a casual watercooler... and long time overclocker... I'd like to express my frustration in simply reading the posts on this thread.

All I want to know is which block is better with my particular setup... and no one here can seem to even agree on how to properly test a block. Everyone seems to think everyone else is wrong, but can't translate their findings into something the others can understand. Frustrating as hell to try and read.

I'm a pretty logical guy. All I want to know is which block performs better on an X2 4400 processor with the IHS on. ... as most people I know don't take a damn knife to their new $600 processors.

If you have:
same processor
same motherboard
same ram
same pump
same rad
same EVERYTHING

You simply hook up the storm, mount it and test it 5 different times to get an average. Hook up the apogee, mount it and test it 5 different times... and you will see which block performs better on that particular setup... and would spare us the thermophile physics lessons.

If I had a few temp probes that were worth a crap... I'd do it myself!

BTW... the reason for my interest, is I am looking to upgrade from my WhiteWater block. It seems to be showing its age.

I firmly believe the "what" that you are trying to cool... is the primary factor in choosing what you decide to use to cool it.

Back when I had my XP2500+ running at 2.6ghz... the LRWW was the obvious choice. Now, I have a processor with an IHS, and TWO cores that are both larger than the XP2500 I used to have.

My current 1/2" ID loop consists of:
LRWW
2x BIX 120's (both with fans in push/pull)
maze4 gpu block
dual bayres
Laing D4 Pump

Current setup:
X2 4400+ @ 2.7ghz - 1.65vcore
Load temp as reported by Everest Home - 44c

Load achieved by running dual Prime95 + SuperPi 32m sumultaniously.

ie - my main concern for following this thread... is choosing which block to upgrade to.

to storm... or not to storm. ... that is the question.

thats the point, certain people believe there is no proper proof yet as to which is better, and others believe that the apogee is better. Personally, I dont know what to think yet, i just want to see some actual proof, much like you do.

At the end of the day, there will likely be a 1 degree or so swing between the blocks, so you can probably base your decision off that - does 1 degree matter to you? if it does, then wait. If not, just get whichever tickles your fancy.

lol
to whom do you refer ?
you are quick to damn a company that, to my knowledge, has never misrepresented a product
and is almost alone in providing meaningful product performance data

Orkan
Swiftech has the only data, which many question for mostly junk reasons - no facts
try to distinguish between those who test, and those who wish to direct how testing is to be done w/o paying the budget

since you are not popping the lid, I think the Swiftech test method is spot on as that is just the cooling setup you have to deal with

procooling is morphing to proconfusion because someone's baby has been called ugly by comparison ?

yeah I am in the same gray area you are.... From what I can tell from this thread is that no one thinks that apogee can really out perform the storm and thus nitpicking the testing methods. Personally I can only see one design flaw in the storm and that is the effective cooling area. Being a Peltier man myself I need a larger cooling area for my peltiers making the apogee more for me. Although I do believe that if I were to hook Apogee to a cpu and the storm to a cpu the storm would come out ahead. My simple answer is the storm will do better if you cooling setup can handle it (high GPH/Head rateing pump). The more complex answer is that the apogee might do better for a larger cooling area and when you dont have a 300 GPH/10' head pump. Simple? any glareing errors?

so basically... with a IHS 4400+... and a Laing D4 pump... which only flows like 180gph, with a 6' head... 2 BIX's, and a maze4 gpu block, the apogee would probably be the better choice?

Bill, I am not saying that you or Swiftech, since clearly with your involvement with the TTV you're seem to be taking this very personally, are lying or misrepresenting.

I am only saying that the issue lies with the TTV, and that's ALL that I am saying. If people want to read it as me accusing others of intentionally misleading, then that is their problem. That is NOT what I am saying.

People used to believe the world was flat based upon the evidence they had. There was no intent by those people to mislead. The world is as round as the TTV is flawed.

well, mabye I had not gotten my hands on a apogee yet (will soon though) so I cant give you any numbers. My best bet is in the real world the storm would do better and has been proven to have as low as .05 c/w. So with your setup I would go will the storm. One small word of advice, I am guessing that some reviews are going to come out by the end of thanksgiving weekend and they will have more solid numbers.