• A Great friend to the HardForum with a great kid that he is trying to get a scholorship to continue his schooling. Please give hime a vote! Only 24 hours left! Thanks.
    If you have an VOTE FOR KEENAN!

33-Way Thermal Interface Material Comparison

Seriously, you have no idea how miserable this experience has been.

It's damn near impossible to test 33 different TIM's scientifically, since one or two "require" extended cure times. The other TIM's may not require it, but the manufacturers of at least six suggest it. Ultimately it appears as though we might have to revisit the subject with a one-hour constant thermal-cycle cure time.

The article was easy to write, but the testing has been exhausting. I can see why more sites do not spend the time and effort to cover this topic is great detail.
 
Thanks for taking the time to address this element of system building that mostly has been documented by word of mouth and "a vs b" tests. Really appreciate this...
 
and no mention of type 44 which is as good if not better then all those.and comes in a 1oz tube which will last you for a very long time.:)
 
Good effort, not enough info in the article about ensuring the testing process was uniform, my first question would be, after popping in and removing the stock heatsink 10+ times is there any wear on the PCB mounting holes that would allow the heatsink to cpu claming force to change, ditto on the retention pins etc. I would have preferred a spring loaded heatsink retention system with precision springs where the tighting tourqe on the retention nuts could be measured and thus ensure uniform clamping pressure throughout the experiment. Applicaion methodology was also not mentioned. But I am insane. I also am pretty sure the CPU DTS has at best an accuracy of +/1 C. Since you apparently use the same cpu for all compounds tested that should be the same but no mention was made of any calibaration testing of this factor in regard to drift over time allowing final temps to be off by as much as 2C, critical to the temp measurements made and the outcome of the test as the final temps are within this margin of error. However it still is a very usefull experiment and /cheer.


For those interested in more information on TIM. (although no where near the number of products bravely tested above)
http://www.overclockers.com/articles1389/
http://www.overclockers.com/topiclist/index26.asp#THERMAL GREASE
 
Wow, excellent review! I recently tried finding a good comparison of TIM's, but only found much older ones.

I couldn't agree more about the cure cycles for AS5. I have a huge tube of thermalright stuff that came with my Ultra-90 a year or so ago. It looks like regular, ol, cheap white thermal grease. Recently I started using it when testing new processors, etc. Just to find it's a few degrees better then uncured AS5.

It spreads easier, cleans up much easier, no cure time, and it's dirt cheap.

Brings us back to this. Is a $10-20 tube of thermal compound really worth 2 or 3 degrees?

Jodiuh,
You are correct. The Coolabratory stuff totally eats aluminum. It's something I wouldn't bring near my computer.

I wish I saved the link, but I saw a video where someone put it on a piece of billet aluminum. It sizzled, smoked, and ate away at the metal. Nasty stuff IMHO!
 
This article has become the most discussed project Benchmark Reviews has worked on to-date! The primary complaint is that we did not use a break-in period or cure time for any of the products, to which I reply that if we do it for one that we most do it for all. Since AC5 requires 200 hours, it makes the project impossible for 33 different compounds.

The next complaint was that we didn't include every product known to man. Well, sadly, we didn't. I have never heard of "Type 44" but you can bet that I will track it down now. Coollaboratory is a difficult subject, since it damages metal and can dissolve aluminum. Thanks to HardOCP's own Kyle Bennet, we will also be including the MG 8610 TIM he uses.

I will go back and add some detail to the article, but for reference I used six different brand new Intel CPU coolers during the course of our testing. Hopefully that will suffice for your satisfy your concern BillParish.

Since so much attentioin and controversy has been created, we will revisit the top ten products tested, and add a few new items to the testing. It may sound strange, but someone else actually wanted toothpaste and no material tests conducted. We can't please everyone.
 
I will go back and add some detail to the article, but for reference I used six different brand new Intel CPU coolers during the course of our testing. Hopefully that will suffice for your satisfy your concern BillParish.
Click to expand...

Jeez, I hate to keep sounding negative, I fully realize the amount of work your effort takes and again, I applaud the effort. However it concerns me greatly that it is not realized that introducing 6, even if new, heatsinks into the testing completely invalidates any testing methodology by introducing 6 sets of variables you have no way of measuring and controlling.

I tend to get very picky about details due to my previous life, I will go find someone needing help with their computer.
 
These were all identical heatsinks, and they were used at different points in testing. The first round was tested three times, with a new one for each. The second round tested twice, with two coolers. The third round was only five items, so I used only one.

You can keep picking on my methodology, but I haven't seen you point to a better source or offer positive suggestions to remedy your concerns. Apparently it wasn't good enough when you though that the same cooler was used, and now it's not good enough when you discover it was six. Please, do it your own way, and let me know.
 
Thanks for the article! I know I never would have the patience to do a review like that but I did have the patience to read it all. I now know that copper based grease isn't the way to go. I got a tube of it for free as part of a promotion but won't be using that stuff any longer. Looking forward to reading your follow up review.
 
Reading from the tube:

Type 44
HEAT SINK COMPOUND
NON-CONTAMINATING SYNTHETIC ESTER BASE
EXTRA HIGH EFFICIENCY HEAT TRANSFER​

The stuff works well, although the liquid does separate out in the tube over time, so squish the tube around some b4 opening it if it's been sitting awhile.

Manufactured by GC Electronics (I think), the only place that lists it for sale is directron.com.

HardOCP did a heat sink compound lineup in November of 2001 in which it topped the list in thermal conductivity and also had some other advantages that may make it the thermal compound of choice. In a nutshell, it does not stain like the compounds with significant silver content, and it lasts longer than the silicone based ones. I would also be interested in seeing how the Type 44 would perform against the 33 compounds already tested.

See http://www.hardocp.com/article.html?art=MTQ0LDEsLGhlbnRodXNpYXN0
 
2MuchCoffeeGirl said:
Reading from the tube:

Type 44
HEAT SINK COMPOUND
NON-CONTAMINATING SYNTHETIC ESTER BASE
EXTRA HIGH EFFICIENCY HEAT TRANSFER​

The stuff works well, although the liquid does separate out in the tube over time, so squish the tube around some b4 opening it if it's been sitting awhile.

Manufactured by GC Electronics (I think), the only place that lists it for sale is directron.com.

HardOCP did a heat sink compound lineup in November of 2001 in which it topped the list in thermal conductivity and also had some other advantages that may make it the thermal compound of choice. In a nutshell, it does not stain like the compounds with significant silver content, and it lasts longer than the silicone based ones. I would also be interested in seeing how the Type 44 would perform against the 33 compounds already tested.

See http://www.hardocp.com/article.html?art=MTQ0LDEsLGhlbnRodXNpYXN0
Click to expand...

I'm already ahead of you... in fact, I have just recently completed testing on Type 44. Type 29, and two others from GC Electronics. The testing is taking a VERY long time, primarily because the product count is around 55, and I try to keep consistant testing conditions (humidity, ambient temp, etc). A semi-updated list is available here: http://benchmarkreviews.com/index.p...sk=view&id=138&Itemid=62&limit=1&limitstart=7
 
You must log in or register to reply here.
Back
Top