Official ICD7 Thread

I see that we have a few threads where thermal pastes and more specifically ICD7 are discussed. I've done some pretty extensive testing of my own, and highly recommend ICD7 for notebook use.

http://forum.notebookreview.com/showthread.php?t=369732

I've talked to Andrew from Innovative Cooling, and set up a ICD7 Giveaway....with the caveat that people who participate share their experiences and results with ICD7 usage with notebooks.

I've known Andrew for awhile, he's a fantastic guy thats put a lot of time and effort into ICD7.....he has also been a sounding board for various articles I've written about thermal compounds, and different testing scenarios.

So, lets use this thread to discuss the Giveaway, ICD7 questions, and Testing Results........

 

ICD 7 is a great choice for using as a replacement for both the thermal pad on the chipset and the thermal compound covering the CPU. Its pretty thick stuff, and using it, you dont have to use the copper shim with the chipset as this easily fills that gap. AC MX-2 and Tuniq TX-2 are both very good, but there have a much thinner makeup than ICD7.

http://hardwarelogic.com/news/137/AR...008-03-03.html
http://forums.hardwarelogic.com/f213...tml#post153380

Between ICD7 and undervolting the P8400 I've personally seen a 9C drop at idle and 20-25C drop under load (it varies depending on ambient).

Stock:
IDLE..............36C/36C
LOAD.............70C/69C

Replaced Thermal Compound:
IDLE..............31C/31C
LOAD.............58C/56C

Undervolted Processor:
IDLE..............27C/27C
LOAD.............49C/48C


I have no comparison data for the T9900 (As I didn't have the stock thermal pad), but here are my results thus far during stress testing:

Stock:
IDLE..............38C/39C
LOAD.............73C/73C

Undervolted Processor:
IDLE..............27C/27C
LOAD.............61C/62C

 

mate, i have a T8300 with MX-2 and get 27c ide and 45c load.

 

You don't want a thick paste on your cpu or heatsink.. that stuff is meant to be thin, and only fill microgaps.

 

Maybe a little research is in order? At least you MIGHT check out the other thread (where this was discussed in detail, and where others shared theri results as well). I'm just trying to give this its own thread so we can share results.

 

Thats great......but what does that have to do with this thread? I started this thread to get people to participate in the giveaway and testing.....if you would like to try ICD7, please feel free to participate, and contact Innovative Cooling at the email addy above to get a free sample....if not...move along, nothing to see here, mate.

 

comparing a completely different laptop and cpu doesnt really help..

 

Then maybe you should buy some ICD7 and compare the results

 

I'm trying to help them get some volunteers for testing, so we can build a database of results.

The more people that help out the better.....doesnt matter what model you have, what processor you have, etc.....I just want to see peoples results.

 

e-mailed... and hoping i get good results ^^
(it'll be going on a 7811FX)

 

Just remember, its not the thermal compound you guys are used to (as evidenced by the "You don't want a thick paste on your cpu or heatsink.. that stuff is meant to be thin, and only fill microgaps" type of comments). Follow the directions at the Innovative Cooling website, or use this pic as a reference.

 

Hi,
I purchased the ICD24, the ICD7 I think is too little for placement for the northbridge as well. I have posted before with my P7805-u and P9600 CPU.

Under idle there is little difference, This is because there is little heat to transfer off to the cooler mass of the HS. Turn up Prime 95 on both cores is where the paste shines. I previously had MX-2 on the cpu only and now have ICD on both the CPU and replaced the thermal pad of the Northbridge with it as well, the later I highly recomend against as this is a personal and yet unsustantiated test.

I can say for the CPU under load there is 2-3c drop per core and 2-3c drop per TSZx. I am without a doubt a convert. To E-, yes the material interface is thicker but this is unlike other compounds. because diamond is such a great heat conductor it can be alot thicker with better results, especially where thick copper shims are required (The reason for my testing it on the northbridge)........

EDIT; use the pic above, this applies differently than normal paste................

Also you see to the left the before, but right now at idle as I post I see
TZS0 = 40c
TZS1 = 40c
CPU0 = 25c
CPU1 = 29c

 

I would hardly call it unsubstantiated testing. I've been doing this a long time, and know what I'm talking about. Having tested it replacing the thermal pads on video card ICs and seeing a very noticeable improvement in temperatures and overclocking ability, I started testing it with the chipset on the 7811 FX, and it works exceptionally well, as shown by my results, the results of other reviewers and testers, and some members here.

If you use a dab about the size of the one pictured above, its plenty for the space between the heat sink and the cooler. I think many of you are making way too much of the space.......

TAN, whats the ICD24, you have a link for that? Have not heard of or tested that one yet.

 

the 24 carrot, larger bottle/syringe. I say unsubstantiated as our northbrige without the heat pad requires a shim for there to be preasure against the HS. it can not be paste alone as there is no surface to surface contact on its own. without a heat transfer media such as a pad there would be a definate AIR GAP. I have yet to see anywhere this has been tested under those circustances (read as under no surface contact preasure and filling a total void (possibly over 1 mm)).............

EDIT; for the CPU I give a recomendation, I'm only 24 hours in on the Northbridge an will not recomend a dramatic departure on it with out a longer term succesful test. You can't change my mind there, I've been doing this since 1979 so I am well versed on both the good and the bad...................

 

I will be buying some soon (Maybe 2 weeks) when I get a T9900.

 

Thats simply not true. The viscosity of the ICD7 is MUCH thicker/heavier than typical thermal compound. Try it, put a dab like the one picture above on the chipset, clamp it down, and burn it in for an hour or two....then remove the heat sink....you'll see that there is full contact between the chipset, compound, and heat sink.

Honestly, I don't know why I bother with forums anymore......too many experts, too many people spouting off "knowledge" without even having used the product........Its frustrating really. I do this professionally, and wouldn't be recommending something if I didn't see the results during testing. In two threads now I've shared my experience, had other collaborate it, shown results, confirmed those results.......and it never fails there is some guy thats been doing this 70 yrs who knows better....even though he has never really tested it.

I'll leave this to Andrew and anyone else interested in helping him build a database.

Awesome processor, I'm extremely happy with the performance. Between the T9900 and the two new HDD, I'm extremely happy with the P-7811 FX.

 

A Recent observation;

Before the CPU fan always stayed on a real low setting. This while the CPU and TSZ temps hovered at a steady temp. My TSZ now regularly gets to 40c where the fan cuts out when both get to 39c and lets the temp rise to 46c where it kicks in on low again cycling the TSZ to both under 40c again.

The fan seemed to never cycle before, this is on AC under high perfomance Vista 64 and not undervolted............

 

Sorry I am stating my lack of knowledege and experience with the product. And while I agree 100% with your original analysis of the thicker product I only recommend what I see and have experience with. fter a week you can bet I'll pull it apart and take a good hard look. If I like what I see you can bet either way I'll post it.

Now if I thought it was not going to be a succesfull test I would not take my hard earned investment and my main system and risk them. so I havbe a certain amount of good faith in the product. am I happy so far you bet I am. The northbrige (GAP) here is a unique circumstance therefore it shall be treated as such.

99.9% of heat sinks have original contact pressure between the heatsink and IC. even if this is uneven or surfaces are irregular. There is no surface to surface normal contact here so unless under this circumstance there is an already substantiated long term test and methodology that is documented and proven I'll hedge my recommendations until it has proven itself. I will not be responsible for others killing their systems, or recommending a way for them to do so while voiding their warrenties..............

EDIT; I am not a detractor of the product, I am pleased so far and the northbrige holds alot of promise. I am extreemly hopeful but state my experience as I am no where near a newbie. This is a well proven product under most circumstances. for normal heat paste/pad replacement this is a simple no brainer, and you can bet I'll psot when I pull it apart with pics.................

 

No, its not you.... I think its just frustration on my part. I've counted at least 5 threads dedicated to the topic, and decided I could jump in and share my knowledge and experience......and each time I've encountered the same thing from people who 1. Dont bother researching the subject, 2. Know better, even though they havent used the the porduct in question, 3. Have been doing this since the invent of computers, and know better.

So, excuse my grumpiness. I have spent the better part of 6 yrs writing articles, testing, and working with companies....I'm not talking about a guy sitting at a desk using HWMonitor and Orthos, I'm talking expensive professional testing equipment for everything I do.

Something I think deserves some mention here....ICD7 works because there is no silicon, there is no liquid....so there is no bleed out or loss of composition. As to the gap, ~1mm is an issue when you are talking typical thermal compounds that include liquid or silicon, as when they heat up they spread out, or thin out......thats not a problem here.

C'mon man.....whats the general theme of this forum? To solve problems, or most often to help people make modifications to their systems to improve performance....pretty much any modificatio0n made is going to void the warranty, and I think people understand that.

 

I understand and am with you 100% and agree 100% in its proven applications. I absolutely love this product there for the CPU and GPU, although I have not tried the GPU the product is well proven in this type application with our systems. I agree with the gap situation, the product holds the most promise of anything yet I've seen. There is no where I could find in those other threads where our situation was encountered specifically so while it should work I have to see the results to say "The solution is here people, get it while its hot to make you PC run like its always in the shade!"

 

Thats the whole point of this thread....to hopefully get some experienced people to try it and share their results.

I have nothing to do with Innovative Cooling. I'm just someone who saw multiple threads posted about thermal compounds, and alternatives to the stock thermal pad and decided to help out people here with an opportunity to try ICD7 for free, while also helping Innovative Cooling expand into another market.

The biggest obstacle I've come across in trying to get people to try it is the old thermal compound issue where people cant get past the application of it.

I'd really like to see as many people as possible get involved in this experiment.....The more people that take part, the better chance to confirm the results as a community.

The big issue here is getting enough people to do it, getting them to apply it properly, and getting them to share pics, their experience, and test results.

 

Now people remember something ehre. The idle temp may not show any major difference. The fan kicks on at a certain temp and off at another. The paste won't change this, what it will change is the fan will run shorter and the CPU may stay cool longer meaning longer intervals between the fan kicking on.

Your true measurable results will be at high CPU load. As I've stated I see 2-3c less than with the MX-2 that was previously applied............

 

I will email you tomorrow. I build pcs all the time at work and can test this stuff on all kinds of processors and gpus, definitely want some for my personal p7801 laptop

The latest pc I built at work:


If there are good results we will submit an order to purchase more

 

I just downloaded hwmonitor, and my cpu reads, Core#0 37*C , Core#1 31-34*C idle. My core#0 stayed the same, but my core#1 is actually keep changing from 31*C - 34*C. Is this normal?

 

For desktops, check out the Innovative Cooling site, those graphs are independent reviews, and there are links to several different forums where they have done giveaways/testing like I'm trying to get organized here.

Sometimes, yeah. sometimes the processors have a stuck or bad temp sensor, happened to me on a couple INTEL desktop processors. There are several variables that come into play here (load on the cores, etc).....What processor do you have? Are you undervolting it? What temps do you see under load?

 

the whole point i made is that i have a cpu with a higher wattage, and it is cooler in a 14.1" laptop than his P8400, and besides i won't want to trouble him with shipping to the UK

 

Well i have used the IC7 in Desktop PC (on Q6600 v.B3 and on 4850 GPU), now on my P7811-FX. The results are worth mentioning. I have used AS5 on the same Desktop system, and the temp was higher with AS5 than IC7. Also i have applied the IC7 to my P7811FX CPU,GPU and NB. And in every case the temperatures got better. So end of discussion here, if IC7 is suitable for p7811FX northbridge or not. It is. I have 2887 model of P7811FX. But i do no this that P7811FX had 2 different types of heatsinks in. Some had bigger copper plates, and other smaller. If i remember correctly. I have the one with that bigger copper plate, which is weird because the bigger plates were common to later release of P7811FX (version 2888?). Maybe that is worth mentioning, if that IC7 doesn't fill the gap for them, maybe it's because different heatsink. My heatsink and NB gap was about 1mm for sure.

 

How does this stuff compare to OCZ Freeze?

 

I know that OCZ freeze is little better compared purly the temp. But i dont know how suitable the OCZ is for NB. Is it thin paste or thick like IC7?

 

I have many notebooks here to try it on too. I will send the email with pics later today to see if I 'qualify' for a sample

 

OCZ offers the absolute best performance in all of my testing, but the top 4-5 are all lumped together temp wise. Freeze has a very thin consistency, much like MX-2 and TX-2, so I wouldnt recommend it for the NB chipset as it will bleed, causing loss of surface contact.

OCZ Freeze
ICD7
Arctic Cooling MX-2
Tuniq TX-2
Shin Etsu G-751

It just depends on the application, price, and personal preference. The reason I've been recommending the ICD7 here is the chipset. The consistency/viscosity make it a good replacement for the thermal pad, and a possible alternative to the copper mod.

I'd like to keep this thread on topic, and round up some volunteers for testing.

 

yea im in for testing, my laptop should be back anytime, finally, my X7800 needs all the help it can get.

 

First i want to ty for responding to my question, i appreciate it. Hmm under load, its about 45*C on core1 and for core0 it started moving up also, it wasnt stuck anymore at 37*C. When i get the chance ill take a screenshot. My CPU is the stock one that cme with p7805u. P8400. No undervolting, i downloaded hwmonitor right away right when i got home from bestbuy. Id love to test that id7 also as i already ordered my p9600.

 

From what I've seen of eviews the OCZ Freeze does not to seem to rate better than MX-2. This is the reason I never went that route. So far again I am happy with the ICD............

 

Like I've said, there is a group bunched at the top...usually ICD7 and OCZ Freeze at the top, then a group a bit behind them (TX-2, MX-2, Shin Etsu) then the bulk somewhere in the middle (Silver based compounds....then the generics.

I want to touch on the Cool Labs Liquid Netal......it always shows up at the top, or near the top, but its extremely corrosive with aluminum, and from the chem people I talk to would be corrosive with any metal over a period of time.......its also a to apply and to remove, and comes in a neat medical syringe complete with a really sharp needle.

 

OK you win!!! gimme some

 

How does the ICD24 compare to ICD7? I don't see much about it on their website.. ICD7 is 7 carat and 94% pure diamond by weight and ICD24 is 24 carat and 92% pure diamond by weight.

 

The ICD24 is just a larger quantity...............

 

Looking forward to applying it tomorrow after lecture!
Now if I can just find a way to get to my gpu...

(Will edit the post with results)

 

This definitely looks like its worth a try. I'll be spending my summer in Taiwan so I'll need all the cooling I can get. I know that the p7811 is similar to my p7805u but anyone try it on the latter yet? Also anyone have experience with applying this to the gpu? That seems to be my biggest concern in heat factor as underload it hits 85C even underclocked.

 

Went was the last time you cleaned out the heat sinks?

The one for the gpu should be in the upper left hand corner- that could be your largest problem.

But, from what I've seen, basically no one has gotten to the gpu. These machines seem to be build is such a way that the technicians laugh themselves to sleep at night at the prospect of us taking it apart.

Maniacally laughing.

 

Actually, it isnt that difficult, now when i have taken it apart at least 3 times. I can say its easy, only you have to pay attention to details. And ofcouse take your time. GL

 

First image is before the thermal paste.
Second image is after the thermal paste.

As you can see, I have a 3C drop in 1 processor, and 4C drop in the other.

Still trying to get to the gpu, so one i get it I'll post that up.

Also, I was curious, is it ok if the paste is also on the larger part of the chip?

 

its ok because its non-corrosive and non-conductive, but its something you really dont want if you can at all avoid it.

 

So what's the general consensus on this ICD7 stuff?

Does it or does it not replace the need for a copper shim once the thermal pad is removed?

 

So far so good here, I still haven't had the time to pull it appart but can say running it there have been no issues. Get the larger 24 and be generous on the northbridge, twice the amount of the cpu is good to be sure the gap gets filled and the chip is still fully covered...........

 

so i did the compound switch and... O_O!
i also cleaned-out the heat sink while i had the lappy opened
(so my after temps are heat-sink cleaned-out + icd7)

Before-idle
Before + 15 mins of gears of war
After-Idle
After + 15 mins of gears of war
what a diff in temps that made!!! (i only aplied it on the cpu tho...) also notice how it even made my wallpaper even more awesome!

EDIT: forgot to say... this was on a Gateway P-7811FX stock only added another 500gb HDD and icd7 now ^^

 

Still wondering if anyone has able to reapply grease to the GPU?

Deron: My computer is less than 3 weeks old... I hope my house isn't dusty enough to be blocking the fan already!

 

Well, my venting hole on GPU fan got extreamly dusty with around 3 months. And i clean my house regulary in a week. But i benefited from paste change on GPU.

 

is there a guide somewhere on these forums for applying to GPU? all the ones ive seen are CPU.