To De-lid or Not to......

[Silver_Pharaoh]

Getting a lot of nah's and no's. I'm actually kinda shocked!!

Thought there would be some heavy peer pressure here.....

Not on a chip this expensive!

 

[RGone]

Getting a lot of nah's and no's. I'm actually kinda shocked!!

Thought there would be some heavy peer pressure here.....

Well I don't think I indicated either direction other than I did not think you would fail twice in a row on FX-9590. I know at one time you were working hard at two jobs and about to work your arse off and we know because it seems to have a 'crack' in it. Hehehe. So I just sort of thought it better maybe not to attempt something iffy if you just did not have to have lower temps. Right now you have a running rig and pretty stout at that. As long as it is working fine now...well you get my drift.

RGone...skisterzzzzzzzzzzzzz

 

[MaddMutt]

I would have to go with your second post

Perhaps waiting till it's so dated.......

You are working 60+ hours a week..... Been there

You will not get the "Nice & Tingles" with that many hours under your belt..... I Know

Wait until you are on a vacation or can schedule some days together to plan and enjoy the process. It REALLY SUCKS when you are deep into the "MOD" and have to stop because of something else
Because you had to stop, that will be playing through your mind, while you are suppose to be doing something else.
If something does go wrong with the "MOD" and you had to stop, that will also be going through your mind > Been there

 

[RGE]

OK so your saying re-using the IHS plate makes only maybe 5c difference at best. Ya I can see that....

The point of why I de-lid is to not use the IHS plate at all. Why spread heat, I wanna transfer it before it's spread.

A later thought comment:



Replace Solder with TIM to directly contact the waterblock. Diamond particles measuring 0.0000015 cm rated at 8.3 w/mK or something of the sort within the paste.

Now removing that plate that seems to accumulate heat, then transfer it to the water block seems to make for a slow process.

Since Copper and Silver are really close together for conducting heat, It doesn't matter which is used. temps don't drop enough to replace copper plate with silver. But every little bit helps, so along the way, tried it.

In theory the IHS plate isn't really needed. It's dummy proofing really.

In the end solder is garbage for conducting heat. Taking IHS plates off AMD chips has done nothing but helped.

No, im saying removing an intel soldered IHS and going bare die with liquid metal will end with roughly the same temps, up to max 5c better, doubt AMD is much different. Removing solder and going bare die with paste you will see worse temps, ie you are not accounting for surface area. I tested this long ago with intel i950, as did ?fallwind on extreme, many years ago. He even tried water directly to die, which told him up front that wouldnt work.. Multiple people now have tried bare die with liquid metal vs ihs back on with liquid metal the temp difference is 4-5C with IHS on or off, if liquid metal used for both. liquid metal is 40 w/mk. intel solder is 87 w/mk.

Intels solder conducts heat at 87 w/mk. best paste is about 5 w/mk when tested in independent labs, but whether 5 or 8...at 8 w/mk that is 11x worse than intels solder....I linked the specs above.

So where do you see solder is bad a conducting heat? Are you saying intel is wrong in posting there spec of 87 w/mk per link/quote from intel paper above?

 

[ShrimpBrime]

No, im saying removing an intel soldered IHS and going bare die with liquid metal will end with roughly the same temps, up to max 5c better, doubt AMD is much different. Removing solder and going bare die with paste you will see worse temps, ie you are not accounting for surface area. I tested this long ago with intel i950, as did ?fallwind on extreme, many years ago. He even tried water directly to die, which told him up front that wouldnt work.. Multiple people now have tried bare die with liquid metal vs ihs back on with liquid metal the temp difference is 4-5C with IHS on or off, if liquid metal used for both. liquid metal is 40 w/mk. intel solder is 87 w/mk.

Intels solder conducts heat at 87 w/mk. best paste is about 5 w/mk when tested in independent labs, but whether 5 or 8...at 8 w/mk that is 11x worse than intels solder....I linked the specs above.

So where do you see solder is bad a conducting heat? Are you saying intel is wrong in posting there spec of 87 w/mk per link/quote from intel paper above?

OK fair enough....

So RGone has this 4300-FX. I've already de-lidded it. He's gonna send back at his leisure.

I was hoping my friend to do a write up, but Life happens and he's not gotten to it. No biggie.

Gladly give you a chance with it for a write up.

On the spec notes.....

35.3 W·m−1·K−1 // Is the conductivity of Lead the basis of which solder is made. It is very unlikely that their specs are spot on. Lead conducts very poorly. So how they more than doubled Leads ability to conduct heat is amazing, I'm not saying this could be done.... I'm just curious how?!

Copper is 400'ish. OK that's good..... But you have to heat and cool that IHS plate. THIS is not instantaneous in any way shape or form. Why....

First core heats, then solder, then IHS plate, then thermal paste, Then waterblock, then water.... rads, pump even the tubing warms up. And it's all supposed to....

BUT the first 4 or 5 listed in the sentence above are important. WHY??

Simply because the faster the heat hits a waterblock, the faster it's removed from it's source aka cpu cores.

So my de-lidded cpu transfers from cpu to TIM to waterblock, then water, then rads, then air.

Now, correct me if I'm wrong, but wouldn't transfer of heat via waterblock to die, be a faster removal of heat? Why would solder be better than a very thin layer of TIM?

Also the mention of surface area. OK. It's designed for air cooling. sA didn't have IHS plates. The chips worked great.

I tested this long ago with intel i950, as did ?fallwind on extreme, many years ago. He even tried water directly to die, which told him up front that wouldnt work.. Multiple people now have tried bare die with liquid metal vs ihs back on with liquid metal the temp difference is 4-5C with IHS on or off, if liquid metal used for both. liquid metal is 40 w/mk. intel solder is 87 w/mk.

Ahhhh it wouldn't work because of mounting pressure I assure you. If the PCB didn't bend, it wasn't tight enough. I get mine so tight, the board warps enough to create a no post, meaning pins where not touching contacts (AM3/+). Evenly back them off half a turn and BAM, posts right up. So a good tight EVEN mount is required. That is all I can point for this direction as results and cpus will vary.

best paste is about 5 w/mk when tested in independent labs, but whether 5 or 8...at 8 w/mk that is 11x worse than intels solder....I linked the specs above.

Directly speaking, the Antec F7 paste.... Diamond particles measuring 0.0000015 cm rated at 8.3 w/mK.... is speaking of the particles, not the actual paste. If it where directed at the paste, a drop of ketchup would be the same thing.

Since diamonds conduct heat better than almost every other material out there, HOW DID THEY COME UP WITH SUCH A SILLY RATING? Diamonds are 1000 w/mk.

I'm saying there's something missing. Last time I checked I got a good temp drop just from going from Artic 99.9% silver pieces to 99.9% diamond pieces.

Maybe it's each cm = 8.3 w/mK perhaps? (surface area) ---- Something that's soldered/welded no longer has a surface area. It's perma-bonded.... at least till I get hold of one.

But again, If you'd like to run the FX-4300 with and without the IHS plate, I'd gladly send it your way for testing. You may be in for a bit of a shock honestly.

 

[MaddMutt]

I'm not sure on this.... I read somewhere, on here, OCN, or it might have been (H), I'm not sure.
Anyway, the talk was about Intel's IHS and how after awhile the Inside TIM no longer contacted the IHS. Air is a POOR conductor of heat.
They went on to talk about de-lidding the CPU, to help lower temps = higher over clock.
Is this what you are referring to on the 2600/3700??????

 

[RGE]

Intel uses indium solder. The conductance of indium is 82 w/mk. You cant just use any solder on a chip, and no one has used leaded solder on cpus in many years. There is an entire corp called Indium via intel that sells indium solder meant for cpus... has to be soft, correct melting temp, weather thermal cycling without voids or damaging chip, etc.

Any intel cpu that uses paste, because of the thick bondline and poor 5 w/mk intels polymer tim conductance, yes you get big temp drops. if you delid for example intel 4790k, and substitute 60 micron thick 5 w/mk paste for 10 micron thick 40 w/mk liquid metal, then temps can get 20C lower. If you then go bare die with liquid metal, you can decrease temps another 4-5C. There are hundreds of people including myself (I have posted my temp drop) that delidded 3770k/4770k/4790k and subtituted intels very thick 5 w/mk tim for a very thin 40 w/mk...that is the big temp drop...but whole reason everyone started delidding is because intel quit using indium solder 87 w/mk, and using much worse 5 w/mk paste.

If there was an air gap, the processor would throttle at any load at stock....but that isnt the issue. the issue is the thick bondline caused by the adhesive, and question if partly design to minimize thermal cycling tim fatigue.

But solder is 87 w/mk, and whole point of delidding is because intel quit using the better solder. I would take a soldered chip over a delidded chip anyday, as would 99% of those that delidded their 4790k/4770K....should see all the threads complaining about intel no soldering the mainstream chips....hence the paste...hence the delidding craze.

I have run bare die before, have pics somewhere on xtreme ? 8 yrs ago. you have to remove the socket and easy to get uneven pressure that causes buggy ram channels. Pain in the neck for an additional 5C gain over using IHS with liquid metal. If mobo sockets and cpus came ready made for delids, and chips I want had paste, not solder, and socket retention mechanisms were in place to help with uneven pressure...then no problem with bare die....but too much hassle, too little benefit. But delidding paste chips, yep those I would if temp limited.

But to each their own. If you want to delid just for testing, im all for it, the entire reason I did it. Just wouldnt expect much temp gains from bare die vs solder. And if you got a big temp drop with solder, that would only be possible if solder joint had failed...ie huge voids.

 

[chrisjames61]

um... brand new 9590 is $208... cheapest 8 core fx i see is $130... doesnt sound like selling the one to buy two is really that possible...

um, the op lives in the USA where everything electronics related is a good bit cheaper than in Canada. In the states you can get the 95 watt 8310 for about $100 as they frequently go on sale. And the tray edition 9590 is $234.99 not $208.

 

[dyckah]

um, the op lives in the USA where everything electronics related is a good bit cheaper than in Canada. In the states you can get the 95 watt 8310 for about $100 as they frequently go on sale. And the tray edition 9590 is $234.99 not $208.

i used pcpartpicker on USA to get those prices...
in canada, 9590 is $280, and cheap 8 core fx is around $160

 

[MaddMutt]

Intel uses indium solder. The conductance of indium is 82 w/mk. You cant just use any solder on a chip, and no one has used leaded solder on cpus in many years. There is an entire corp called Indium via intel that sells indium solder meant for cpus... has to be soft, correct melting temp, weather thermal cycling without voids or damaging chip, etc.

Any intel cpu that uses paste, because of the thick bondline and poor 5 w/mk intels polymer tim conductance, yes you get big temp drops. if you delid for example intel 4790k, and substitute 60 micron thick 5 w/mk paste for 10 micron thick 40 w/mk liquid metal, then temps can get 20C lower. If you then go bare die with liquid metal, you can decrease temps another 4-5C. There are hundreds of people including myself (I have posted my temp drop) that delidded 3770k/4770k/4790k and subtituted intels very thick 5 w/mk tim for a very thin 40 w/mk...that is the big temp drop...but whole reason everyone started delidding is because intel quit using indium solder 87 w/mk, and using much worse 5 w/mk paste.

If there was an air gap, the processor would throttle at any load at stock....but that isnt the issue. the issue is the thick bondline caused by the adhesive, and question if partly design to minimize thermal cycling tim fatigue.

But solder is 87 w/mk, and whole point of delidding is because intel quit using the better solder. I would take a soldered chip over a delidded chip anyday, as would 99% of those that delidded their 4790k/4770K....should see all the threads complaining about intel no soldering the mainstream chips....hence the paste...hence the delidding craze.

I have run bare die before, have pics somewhere on xtreme ? 8 yrs ago. you have to remove the socket and easy to get uneven pressure that causes buggy ram channels. Pain in the neck for an additional 5C gain over using IHS with liquid metal. If mobo sockets and cpus came ready made for delids, and chips I want had paste, not solder, and socket retention mechanisms were in place to help with uneven pressure...then no problem with bare die....but too much hassle, too little benefit. But delidding paste chips, yep those I would if temp limited.

But to each their own. If you want to delid just for testing, im all for it, the entire reason I did it. Just wouldnt expect much temp gains from bare die vs solder. And if you got a big temp drop with solder, that would only be possible if solder joint had failed...ie huge voids.

Do you know if AMD suffers from the same problem of using lower quality TIM inside the IHS????
I know.... The only way to find out is to de-lid a AMD FX-43xx/63xx or some 8320/8350 CPU's.

I never looked at that, as to me this was for Sub-Zero benching only and not water cooling

 

[RGE]

Any cpu that has paste tim will be around 5 w/mk, just arent any much higher ones used in tim1. But I dont know which amd cpus use paste vs solder for tim1, recently I have been more intel. my guess is if you search a particular cpu for delid...someone will have done so and can see

 

[ShrimpBrime]

Any cpu that has paste tim will be around 5 w/mk, just arent any much higher ones used in tim1. But I dont know which amd cpus use paste vs solder for tim1, recently I have been more intel. my guess is if you search a particular cpu for delid...someone will have done so and can see

None. ALL Phenom I and Phenom II and FX chips are soldered including AM2 6400+

The list of tested cpu's that ShrimpBrime has de-lidded is as follows with full temp testing..... All these are soldered chips...

Athlon 6400+ Mr. Scott and Shrimp tested
Athlon 6400+ #2
9850BE
9950BE
970T _____ customer on this forum air cooled gained 100mhz 4ghz to 4.1ghz
1090T
965BE
940BE
FX-5000
FX-4100 x2 (killed one)
FX-4300
FX-6100 IMOG won at CP contest
FX-8320 Mr Scott tested and daily rig for some time. Over 5ghz chip NP. I think he was easily hitting 5.4ghz (don't quote me though!!)
FX-9590 killed this one bought another

And many many non soldered processors.

If I make a claim there IS a considerable temperature drop, I assure you it's true. Specs on TIM and solder need not apply.

Do you know if AMD suffers from the same problem of using lower quality TIM inside the IHS????
I know.... The only way to find out is to de-lid a AMD FX-43xx/63xx or some 8320/8350 CPU's.

I never looked at that, as to me this was for Sub-Zero benching only and not water cooling

See list above. All FX chips are soldered. And they all look like 8 cores under the hood.

All water or air cooled as well.

4790k/4770K....should see all the threads complaining about intel no soldering the mainstream chips....hence the paste...hence the delidding craze.

Firstly we shouldn't compare Intel to AMD for heat or processing power. It's just not fair.

I was running 3770K 5ghz on liquid less than 80c at 1.51250v. No de-lid required.

BUT we are talking about much lower wattage chips with a much higher TDP threshold!! Your talking Intel 77w vs AMD 125w, 140w and 220w AMD processor. That's why no need for solder I betcha!

Could you imagine an Intel Chip having a max threshold of only 62c??!!! You think Intel guys cry now hitting 80c.... Fudke man, they'd be crying harder. Maybe even crapping their own pants.

If AMD had 110c threshold, I'd be hitting 6ghz on air.

Oh and to clarify, I'm not arguing TIM is better than Solder. I'm arguing the IHS plate is useless. The solder AMD uses is poor heat conductivity. They take 125w cpu, use at best 87 w/mk and transfer this to a 400 w/mk Plate, then 8 w/mk TIM then to some more 400 w/mk waterblock.

But since solder is a bonded material, TIM is not.

When you bond to metals together, the weakest point is that bond. The Solder on AMD chip with only 87 w/mk at best is crap for dissipating my FX-9590's 220w output. The TIM works better using without the IHS plate. That is basically my full statement and testimony. I've never re-used an IHS plate after I removed it. It's absurd.

So to replace solder with TIM, I'm in full agreement with you rge. Except that one 970T case. That was reused with TIM. Dude gained 100mhz... freak occurrence perhaps.

Also, I've always wanted a pure silver or even gold waterblock. Funds stop me from having such luxuries.

 

[RGone]

rge and S_B...

...you two guys have had an interesting back and forth of both ideas and personal results. With all that is written taken into account, I sense a few points that are probably the most valid. For me they were anyway.

#1. Intel dorked their average consumer that thinks to clock up with a sort of TIM and not solder between die and IHS on many of their later processors. Note it is believed "many" and not "all" of their current for sale processors.

#2. I heard clamping pressure mentioned. Yeppers, a solid consideration in my mind. At times the strangest things give us answers.
Okay I bought this air cooler which has been superseded in their lineup by one that actually comes with push/pull fans as their big dog. I had to purchase a second fan to actually go push/pull on air.
Aegir SD128264
Then I added this to my Aegir cooler. And it "was" the difference in being actually able to attain 4.8GHz and a shade more P95 Blend mode stability for at least 2 hours using air cooling. Clamps the "pee" out of the cooler.
Crossbar mounting bracket kit for Xigmatek air coolers.

Still in line with #2 is what my water block hold down looks like. The standard Dtek hold down could n0t live with the springs I was using on my water block. Some part of a computer case yielded up mini-channel which I used to reinforce the area across the block to keep the water block hold down from just collapsing down onto the motherboard as the clamping force was increased.

So continuing on inline with this #2 idea, I don't have any mounts on any of my FX processors that are n0t clamping force adjustable to the 'greater' degree. Clamp-it if you want cool has become a given "for me".

#3. Looking at the 'whole' of the DE-Lid idea, it would appear not so much for the average user. Okay let us make that not for the average AMD FX processor user in todays landscape. Trying to get this to where it means something since there are exceptions to almost everything that man dreams up.

Back in the day. What day was that? This day > AMD Athlon XP 3000+. I had one of those over 3,000Mhz before most did on air. I did it by DE-Lidding when there was no solder. Did it on my DFI mobo before I went to work for them. Probably helped gain that job. Hehehe. Back on track RGone... I used a TT Big Typhoon clamped to hale and back. Would do 8 hours of P95 Blend up there and cannot remember what else but the forum for DFI at the time used P95 Blend (8 hrs) and at least two other criteria to consider a clock as "real". Yessir Real hard arse back then. So here in #3, I see DE-lidding as a goodie and clamping as a 'must'. That much has become quite apparent out of the back and forth between rge and S_B. Good show men.

#4. Wish I had had my circumstances where I 'had already' tested S_B's DE-Lidded FX-4300. I do need to test it or return it. May see what the cheapest FX 8 core I can find and DE-Lid the sucker since I know I can clamp it and some ultra metal or diamond TIM and give it a go. You know just for the hale of it. Be a fun experiement. Costly maybe but often inquiring minds just want to know.

#5. I found this and it would seem to make sense to put copper to the die for the best results. But my copper water block seems not as tough as I might wish so may have to get me something copper to put between my water block and the cpu die and then clamp it. Just doing my own outloud brain-fahting.

Thermal conductivity data, @300 Deg. Kelvin (room temperature) in Watts/Meter-Degree Kelvin.

Silver, Ag - 429 W/M-K
Copper, Cu - 401 W/M-K
Aluminum, Al - 237 W/M-K
Cast Iron, Fe - 80 W/M-K
Hi-Carbon Steel - 35 W/M-K
Stainless Steel - 15 W/M-K
Glass/Ceramic - 1 W/M-K

So it seems I would do better with copper against my cpu die than what seems an aluminum IHS? Humh?

Too many people read a back and forth for the definitive or 'absolute' answer as relates to some situation. i don't think there is an absolute answer to the DE-Lid discussion as there are many variables and not the least of which is DE-Lid failure and dead cpu and then clamping down on a silicon die and not cracking or ruining it. Yes I have made De-Lid on early die types and clamped and cracked the die. So the DE-Lid idea is with many variables and not for the average user. You really have to WANT it to get it and it often costs to play. Hehehe.

Ole faht over and out. Keep up the good work men.

RGone...skister.

 

[RGE]

There is no aluminum ihs, ihs is copper with too thin nickel plate to matter for nickel conductivity.

On an intel cpu, yes best for temps is copper waterblock to bare die using liquid metal at 40 w/mk or better yet solder die to waterblock with 87 w/mk solder (not practical though).

The second best is soldered IHS (only 5C gradient through solder 87 w/mk and copper ihs 400 w/mk), then 5-8 wmk paste tim then waterblock. That is better temps than bare die with paste tim, and this is where surface area confuses people.

The argument is why the extra layer of copper and 2nd tim, just impedes temps with the extra layer. Now use that analogy on other things. The goal is to get the die temps to the air, air is always the final destination (even in water cooling). So take an air cooler with massive surface area. Why not remove the extra air cooler, it is just more copper in the way, and remove ihs...now blow air on the die....your cpu throttles and then shuts off. there is not enough surface area for air at .025 w/mk to absorb heat fast enough....you need massive surface area to transfer heat at the paltry .025 w/mk speed. So what speed is fast enough. Water on bare die, again, i920 by fallwind, throttled at stock speed, why water at .6 w/mk cant absorb heat fast enough at very small surface area of die. So is 5 w/mk fast enough with paste...well it isnt as fast as 87 w/mk solder. The real question is what speed is fast enough on bare die to be better than soldered ihs.

So now that surface area matters, where is the limit. The die has very small hot spots. These small hot spots require spreading to a much larger surface area at a very rapid speed, before running into any speed blocks, especially air. But 5-8 w/mk is a massive speed block on bare die. The power density of intel cpus is very high, ie high power, very small hot spots, smaller than die. Gpus much less prone to surface area issues as much lower power density (power spread over much larger area). You can not put paste on bare die on an intel cpu and have better temps than using solder to spread the heat to 10x larger surface area of ihs (small hot spots vs much larger entire ihs. Intel has multiple white papers on heat dissipation and IHS. If you use liquid metal at 40 w/mk, then yes you get better temps by removing solder ihs. .025 w/mk way too slow, .6 w/mk way too slow, 5 w/mk little too slow, 40 w/mk plenty fast. Why do waterblocks need extremely small and tall pins...again surface area must be massive to cool with air/water.

No question if you have die-- 5 w/mk paste -- ihs --5 w/mk paste--- cooling block, and you remove one layer, die and waterblock, that is a no brainer....better temps with IHS off.

But if you have die--solder 87 w/mk---ihs---5w/mk paste --- cooling block....you can not conclude that removing the solder/ihs and replacing with bare die 5w/mk paste waterblock is better, this ignores surface area, just like the example of removing an air cooler and blowing on die with air ignores surface area. All that extra copper fins/cooler is necessary even though it is in the way of air hitting the die. If you solder the waterblock to die at 87 w/mk....then yes you get better temps obviously. But whether you get better temps with removing solder or not and replacing with bared die and paste, with very small surface area hot spots....5/wmk paste is too slow dissipation at the critical low surface area stage...but that depends on power density. Low power density (means large die/high surface area) sure, with intels high power density...you cant put paste tim on die and get best temps...40 w/mk liquid metal, yes.

 

[Silver_Pharaoh]

I see where you are coming from.
High w/mk on the small die, and lower w/mk TIM on the larger IHS -> CPU block.

Rather than lower w/mk on the smaller die -> CPU block.


First option sounds much more efficient in theory at removing heat, but in reality, it seems quite the opposite for most...
That I can't explain..

 

[RGE]

There are tons of people that have measured it on xtreme using accurate methods years ago, and many on oc.net more recently. If you post there that it is better to delid a soldered die and put paste tim on, you will get pummeled by many that know better, because they have measured it themselves. This is the first forum I have ever heard of anyone saying that putting 5 w/mk paste on hot spots on die cools better than a soldered ihs (because of surface area limitations of hot spots). Even intel themselves state they developed 87 w/mk solder because power density increase made paste 3-4w/mk unsuitable...link above. So guess it depends which forum your on.

If you go bare die, you do liquid metal or you dont go bare die. To be fair, gpus do not work this way, they have massive surface area/low power density in comparison so the difference between 5 w/mk paste and 40 w/mk liquid metal is much smaller. Just like TIM studies show only a few C difference between 40 w/mk liquid metal and 5w/mk paste when testing tim2 (between waterblock and ihs...surface area is large, speed of heat transfer less critical).

And hundreds of delids in the delidding section on Oc.net (2,700 pages of delids)...no one uses paste on die, there is 10-20C difference (higher wattage tested/higher the difference) between liquid metal and paste...again because of very small surface area for heat transfer...the effect of speed of heat transfer is magnified by 10x smaller surface area.

 

[ShrimpBrime]

Spreading heat is not disipating heat. Simple as that.

Might I add we are in the AMD section. Intel results from other forums need not apply.

Ive done my testing with excellent results. The admin of my team will confirm.

Have a nice evening sir.

 

[RGE]

correct. But if you say you dont have to spread the heat (increase surface area) to better dissipate the heat.....then try blowing air directly on the hot spots on die. Pins on waterblock spread heat, so it can be better dissipated....cut those off...what happens?

surface area is critical.

 

[MaddMutt]

@RGone

Not sure if you have any OLD video cards laying around gathering dust. I took a old NVidia 6800GT AGP video card and took the HS off, this thing had a massive 1/4" by something like 6" x 4" off PURE Copper. I sacrificed the card to make several Copper shim's if needed. AMD and NVidia have moved away from this as the cost of Copper has gone up. If I took one of these shims and placed it between the IHS and the Water block, what would I need to make this setup work??? I know with the shim, I can get more pressure on the CPU.
All Ideas are accepted

 

[EarthDog]

Great discussion!


Side note:

Spreading heat is not disipating heat. Simple as that.

http://www.synonym.com/synonyms/dissipate/

Sense 2:
disperse, dissipate, scatter, spread out
separate, part, split

... have a good evening sirs.