Yeesh IC Diamond 7 is some crazy stuff

[ICD7]

CLP;

IC7 will scratch up the CPU IHS and heatsink base just on application using the grain method as someone already mentioned. Removal with a Q-tip soaked in solvent thins it out and you can hear the grit working as an aggressive polish/rubbing compound. My first victim - a golden egg 3520 - became very shiny and the second remount and cleaning thereafter had a near mirror finish without a letter to be found! IC7 is also a warranty invalidator! Technically you are supposed to use the Intel HSF to keep your warranty anyway.

It's been out for a couple of years with no problems and may be a localized issue that's magnified under the the lens of the forum.

To address this, I have a giveaway going on at Tom's Hardware and I will ask them for observations wear and tear. I use the compound every day and do not share your experience. So let's have a hundred or so independent users evaluate it and post their comments. I believe it will be a fun exercise to get down to brass tacks on the issue.

 

[WhoBeDaPlaya]

Just took off my HSF (Xigmatek Dark Knight) for a peek today. No noticeable etching/sanding/marring and I applied it using ICD's recommended method (drop in the middle and clamp down the HSF).
Is it worth the price over Chill Factor? Probably not, but the i7s are the only systems that I used it on and while I still have a couple of syringes left of CF, I wanted to try something different. No AS for me (used to be sponsored by them for HSF evals) - it performs well but is messy as heck to clean up.

 

[Suspicious-Teach8788]

It's been out for a couple of years with no problems and may be a localized issue that's magnified under the the lens of the forum.

To address this, I have a giveaway going on at Tom's Hardware and I will ask them for observations wear and tear. I use the compound every day and do not share your experience. So let's have a hundred or so independent users evaluate it and post their comments. I believe it will be a fun exercise to get down to brass tacks on the issue.

What is the suggested cleaning method for ICD7? I utilized IPA and a wet paper towel... lol. If necessary I can use clean wipes although I'll have to jack a few from work

Just took off my HSF (Xigmatek Dark Knight) for a peek today. No noticeable etching/sanding/marring and I applied it using ICD's recommended method (drop in the middle and clamp down the HSF).
Is it worth the price over Chill Factor? Probably not, but the i7s are the only systems that I used it on and while I still have a couple of syringes left of CF, I wanted to try something different. No AS for me (used to be sponsored by them for HSF evals) - it performs well but is messy as heck to clean up.

Did you clean off the ICD7? Is it still greyish? My CPU was all stained from it. Maybe if Ic leaned it better I could read the scribed text better (not that it was completely unreadable or anything)

 

[Markfw]

I tried this years ago, and didn't remember a problem. If I get my hands on some more, I will try it again, and let you guys know my results.

 

[nyker96]

this is news to me, haven't heard anyone complaining here about TIM etching, or scrap off lettering from IHS in one application! that must take quite a bit of abrasive force like in sanding. I think I have a tube of IC7 somewhere haven't used it since I bought it 3-4 years ago, might try it on a chipset first see if it rubs anything off.

So how many here who used IC7 has this problem??

 

[DrMrLordX]

g to happen since welding technically is the fusing of metals together to make one piece. I'm sure with both surfaces lapped you could get a decent fillet (think solder) that would hold it somewhat but when warm it will give easily.

I guess welding is an overstatement then.

Installation and power up with this dust on the cpu is a good way to KILL your cpu!

Tell me about it. That's the last time I lap a CPU without a foam block covering the pins.

Interesting, these exotic TIMs sound like they're more trouble than their worth. Guess I'll stick with the old school goop.

It should be noted that there is some "old school goop" out there that is better than others. IC Diamond 7 technically falls into that category, though the etching . . . well for some its a problem, for others not.

I still swear by x23-7783D (not 7762). It's very easy to use and is excellent, especially if you lap.

If you plan on leaving it alone for a while and allow for proper preparation CLP cannot be bested. It's that good. Muck up the installation, though and I guarantee you will be cursing like the day you get a flat and find out your spare has no air as well!

Some claim that Indigo Xtreme is better on surfaces lapped to 1600 grit. I have my doubts but I have no data to refute the claim. IX is good stuff.

 

[v12v12]

It's been out for a couple of years with no problems and may be a localized issue that's magnified under the the lens of the forum.

To address this, I have a giveaway going on at Tom's Hardware and I will ask them for observations wear and tear. I use the compound every day and do not share your experience. So let's have a hundred or so independent users evaluate it and post their comments. I believe it will be a fun exercise to get down to brass tacks on the issue.

I'm going to dissent regarding this issue; I have been using IC7 since the very 1st release of it, and I've changed out many CPUs with IHS and with my own custom made copper IHS and have NEVER EVER seen any etching on the IHS nor the CPU core. All the lettering is intact and fine. Every system/change was a high OC'er and still no problems... I've had IC7 on my old school XP-M2500, through many CPU swaps, for YEARS and it's never had problems. I might take off the HSF and check just to make sure, but again I do NOT share this problem with the alleged etching...

I will attest to the price/gram though; it sure does "etch" your wallet though lol! I've still got some spare AS3 & AS5 left over, so when the IC7 runs out, I'll prob just go back to AS* since it's cheaper and 1-2C doesn't make much diff these days to me.

Q: How come there's no pictures from the OP: EVERYONE has a camera on their phone...

 

[Suspicious-Teach8788]

I'm going to dissent regarding this issue; I have been using IC7 since the very 1st release of it, and I've changed out many CPUs with IHS and with my own custom made copper IHS and have NEVER EVER seen any etching on the IHS nor the CPU core. All the lettering is intact and fine. Every system/change was a high OC'er and still no problems... I've had IC7 on my old school XP-M2500, through many CPU swaps, for YEARS and it's never had problems. I might take off the HSF and check just to make sure, but again I do NOT share this problem with the alleged etching...

I will attest to the price/gram though; it sure does "etch" your wallet though lol! I've still got some spare AS3 & AS5 left over, so when the IC7 runs out, I'll prob just go back to AS* since it's cheaper and 1-2C doesn't make much diff these days to me.

Q: How come there's no pictures from the OP: EVERYONE has a camera on their phone...

Yeah I got a few dSLR shots of my system too and the clean CPU before putting it in, but I exchanged the CPU. My new one only has AS5 on it as I'm afraid of getting more IC7 on it, but I can certainly try again

 

[JEDIYoda]

What is the suggested cleaning method for ICD7?

Get a high velocity spray nozzle and spray the crap out of it......

 

[lopri]

So how many here who used IC7 has this problem??

I did a while back. Scary thing is that I was actually thinking about using this stuff on a IHS-less core! (Before I learned what it did to IHS, of course)

 

[Rubycon]

I'm just telling my experience.
Perhaps I'm too aggressive when cleaning up TIM? I prefer to remove ALL traces especially if changing compounds. I'm not satisfied until I cannot see ANY traces of TIM that was applied previously - all inspected though a 10X jeweler's loupe.

If the compound is very tacky/sticky/viscous this means some scrubbing may be involved with a q-tip. Solvents are also used to soften/dissolve the base so it's easier to remove. With IC7 I noticed it was quite gritty when thinned out with solvent. It's a good polish I'll give it that!

It's been out for a couple of years with no problems and may be a localized issue that's magnified under the the lens of the forum.

To address this, I have a giveaway going on at Tom's Hardware and I will ask them for observations wear and tear. I use the compound every day and do not share your experience. So let's have a hundred or so independent users evaluate it and post their comments. I believe it will be a fun exercise to get down to brass tacks on the issue.

 

[aigomorla]

I believe it will be a fun exercise to get down to brass tacks on the issue.

then lemme be the first to show u documented on the forum.

This what the op and i both ran into, as well as others:

And the only way to get rid of that is by lapping the processor.

I get no performance hit, and i actually use it as a marker to let me know which cpu is my 975.

 

[Rubycon]

What have you tried removing that with?

Carb and choke cleaner usually remove deeply "stained" parts.

then lemme be the first to show u documented on the forum.

This what the op and i both ran into, as well as others:



And the only way to get rid of that is by lapping the processor.

I get no performance hit, and i actually use it as a marker to let me know which cpu is my 975.

 

[aigomorla]

What have you tried removing that with?

Carb and choke cleaner usually remove deeply "stained" parts.

tried almost everything.. WD40, di-methanol ether as well.

The scrubbing marks are from the brush i used to try to get it off as well.

But no that wont come off no matter what i do unless i lap it.

And i given up trying to clean it.

IF you also read the OP's first post, its the same thing he had a run with.

This stuff etches the IHS pretty nicely. I scrubbed and scrubbed with IPA and it wans't residue. It just simply corroded the crap out of the IHS.

Dont get me wrong... i still use IC7 sometimes.. its great stuff, but just be aware of what your gonna get yourself into when using it, like with liquid pro.

 

[Rubycon]

Why are you covering up your numbers?

 

[aigomorla]

Why are you covering up your numbers?

They are shy.

 

[quadomatic]

I always wanted to try IC Diamond 7. Never got around to it.

If you want something comparable to MX-2 and OCZ Freeze, but WAY easier to spread, try PTI-G3606

 

[Rubycon]

IF you also read the OP's first post, its the same thing he had a run with.

Still looks like abrasion not etching. Etching is a chemical reaction. That CAN happen with CLP. These specialized compounds NEED a specific procedure in both application and clean up. The manufacturers need to supply detailed information (they seldom do!) or a potentially superior product gets dismissed as ineffective or a plain out lie/scam! CLP definitely falls under that category. They do have application videos now, however.

 

[Tsavo]

Why are you covering up your numbers?

Super secret Intel part.

Numbers hidden to protect its identity.

 

[Suspicious-Teach8788]

They are shy.

Looks similar to my CPU. How much did you put on? Small dot? IC7 site says to put a pea size on it. I initially tried the rice grain size of AS5 and I got a small circle imprint when I dismounted.

Then I did the pea sized thing and then it slathered all over the CPU. Then I got a much larger area of discoloration.

Now I'll admit ABRASION can happen when I'm wiping clean with IPA and a paper towel, but I can't imagine that being too severe.

 

[Suspicious-Teach8788]

BTW I did some testing. My MX-3 came in yesterday (Thurs). Since Sat with my new CPU I've been using AS5 24/7. I know that AS5 has some set in time, so AS5 should be more than ok on my system. Well supposedly arctic cooling MX-3 has no need for a burn-in, so I slapped it on. I couldn't find instructions so I used the Pea sized method. I used the rice grain method of AS5 as described on their website.

Temps sucked. I benched 32 minutes of OCCT Linpack and grabbed data. Now perhaps AS5 was at an advantage you say. This is informal testing after all and I had no plan to really make a formal benchmark. So perhaps AS5 is better or perhaps MX-3 could use a burn in. Perhaps the pea sized method is NOT for MX-3.

Well I said fvck it and I slapped on IC Diamond on the CPU after cleaning it. Since this forum is looking for pics I'll provide them after I clean the IC Diamond off sometime in the future. I decided that I'm not returning this CPU or anything so we'll see.

I immediately fired up OCCT Linpack for the IC Diamond and wow. Temps beat AS5 (burnt in), and the fresh MX-3.

Max temps per core:
AS5: 85/82/80/76
MX-3: 87/86/83/81
Diamond: 82/80/79/77 (WINNARRRRRRRRRRR)

Temps are given as the AVERAGE of all 4 cores. Tests were performed with a Noctua NH-D14 with stock fans. Both MX-3 and ICD7 used the pea method (ICD7 recommends that, but MX-3 has no instructions). AS5 used the rice grain method and has clearly burnt in for days. So AS5 may be at an advantage to the two since its had time to set.

But given that IC Diamond 7 was just slapped on like that and it beat out the other two I'm impressed. I left the ICD7 on all night and when I woke up I looked at the Real Temp "high temps" and they all hit around midnight last night. So temps have only gotten better (closer to 75-76 in fact). This shows that ICD7 really whoops AS5. Perhaps MX-3 would do better with an overnight burnin, but do I expect it to drop like 7-8 degrees to catch up? Probably not.

 

[DrMrLordX]

IC Diamond 7 has a fairly long cure time (24+ hours, unknown # of power/load cycles required), longer than even AS5 I think. It will only get better with time.

There are some bad batches of MX3 floating around out there but I can't tell if you got one or not. Near as I can tell, proper use of MX3 requires fairly large amounts of the paste compared to other TIMs. A bead with a 1cm diameter should do the trick.

 

[TheRyuu]

Near as I can tell, proper use of MX3 requires fairly large amounts of the paste compared to other TIMs. A bead with a 1cm diameter should do the trick.

This, from using it and from reading about it on xtreme systems it seems with mx-3 the more you use (within reason) the better your temps. MX-3 SHOULD provide slightly better temps then AS5, but this is all within ~1C of each other anyway so it's not like it really matters that much.

The fact that DLe's MX-3 temps look as much as 3C warmer than AS5 makes me think something's wrong but who knows.

 

[evolucion8]

I used Diamond IC7 and I got slightly better results than with AS5, about 4C cooler, but it indeed took away the CPU etches, but it was because if you scrub it to remove it, it will erase them. I was able to polish my old CPU heat-sink what and it looked almost like a mirror, also I did the same thing with my old pure copper 9500 CPU cooler (Forgot the brand) and looked like new, it is the best stuff for jewelry and shining surfaces as far as you don't erase any marking or etching.

For me, the best TIM is the Shin Etsu which is sold at svc.com, that stuff is awesome and a bit expensive. About 2C cooler than Diamond IC7.

 

[Suspicious-Teach8788]

This, from using it and from reading about it on xtreme systems it seems with mx-3 the more you use (within reason) the better your temps. MX-3 SHOULD provide slightly better temps then AS5, but this is all within ~1C of each other anyway so it's not like it really matters that much.

The fact that DLe's MX-3 temps look as much as 3C warmer than AS5 makes me think something's wrong but who knows.

Probably right. Haha. It was probably a bad mount combined with not using enough. Well it didn't seem like enough when I pulled it off, but I did use a good amount more than AS5. It was closer to a pea size. 1cm bead as someone says is crazy. Maybe I'll try it and see.

But I'm convinced the AS5 and IC7 mounts were good. I've done both of those multiple times and I did get to look at the footprint from AS5 and it looked great.