Starting over on 2600K CPU block thermals, MX-2 Sample Variance?

Posted: March 4, 2012 in Blocks

When my old large 30g tube of MX-2 ran out during my M6 block testing, I was a bit afraid of what that might mean.  I have heard of a few people report variances between thermal paste samples, and I’m finding the same problem here as well.  I had a large tube of MX-2 I started using back in my Q6600 testing round that lasted quite a while through a large group of blocks.  I had been using this same tube of paste as part of my 2600K testing as well and life was good up to that point of running out.  In a failed attempt to “Carry On” with my thermal testing, I purchased a new 20G tube of MX-2 with the hopes that I could simply do some quick validation mounts with my last tested block.

Well, the first few mounts are not looking good to validate the thermal paste consistency.  It’s very clear the new paste consistency has changed between samples, it is visually less dense and more thin in consistency. My old batch 1 30g tube would leave a clear imprint on the heat spreader where the block made good contact, this new 20G tube paste is much different.  It runs as if it was thinned and generally acts differently.

And more importantly, after several mounts of retesting using the new batch 2 sample…

Perhaps TIM pastes can be like a fine wine in that they become more dense and perform better over time in storage, or there has been a slight change in the MX-2 formula, or  there is simply variations between samples out there.  Bottom line for me is that any future CPU block thermal results comparisons will have to start over on testing.  Sorry for further delays on getting caught up on CPU thermals, but with block performances being as close as they are, this thermal paste change and my inability to purchase the exact same thing I was using for previous tests…makes this a stopping point for that round of tests.

I may very well spend a little time evaluating a few other thermal compounds now and decide on something newer/better than MX-2.  I’m not really all that impressed with my new MX-2 sample..it’s just not performing as well as my old 30g tube…:(

I’ve been following Vapor’s work here and now is probably my opportunity to try something different.  If I can’t use my old MX-2 batch 1 results to compare with, I may as well look at switching out the thermal paste entirely:

http://skinneelabs.com/2011-tim-results/2/

Sorry for more delays on future thermal results, but I’m going to have to start over.

Cheers!
Martin

Comments
  1. Ceadderman says:

    Wow that sucks Martin. I’m kinda glad I don’t review blocks. I only do Shin-Etsu G751. I can’t even imagine how many blocks you review or have reviewed but .5g tubes of this stuff is quite expensive in comparison to MX-2. You might try MX-4 which is supposed to be better than the old MX-2 formula. EK went to MX-4. 🙂

  2. PunkNugget says:

    Good review Martin, and good catch on your part on how average this TIM is. I actually used and did a review on this ETI from Enerdyne Solutions – The Indigo Xtreme !!! It’s the BEST out there and NOTHING can touch it !!! I apologize if that’s sounds a bit arrogant, but the facts are what they are and NOTHING does better !!! Watch the vid review to see it for yourself !!!

    But if you’re really looking for the next best TIM and this is a bit too much for you to install and use then use this one:

    Prolimatech PK-1

    Now this TIM is easy to apply and gives you the next best temp rated score (next to the Indigo Xtreme) when you look at the “Average Mount Temperature – Great Contact” pic found on this link here:

    http://skinneelabs.com/2011-tim-results/3/

    for a .5g amount cost me about $15.00 (included the shipping), but to me was completely worth it and I will only have to change out on it only once every 8 to 10 months. With the Indigo Xtreme you only have to change that ETI once every 18 to 24 months. That’s another reason why this ETI is the BEST !!!

    You know Martin, you ought to do your own TIM roundup so we can have something else to compare to with both Martinsliquidlab and Skinneelabs reviews. I think that would be awesome !!! Again, thanks for the review…

    Later, PNugg… 🙂

    PS – I just caught the link review you did on skinneelabs.com – LOL !!! – same review but different page – LOL !!! XD

  3. Martinm210 says:

    Thanks!
    Yeah, I’ve been trying a few of the other top contenders but so far have yet to find one that works better than MX-2 for me on blocks with really good contact. TIM testing results do almost seem to be core specific to some level. I do have this 20G tube so I could always start over with it which should be enough qty for several more blocks. This poor 2600K has probably seen over 100 mounts now, so using a paste that’s relatively easy to use and economical is pretty important in testing. Indigo is a great tool for actual use, but the cost of IX over 100 mounts would be impractical, not to mention the clean up and reflow process each time.

    I’ll probably go ahead and just get a bunch of mounts on the M6 for a new baseline to compare future results. The M6, CPU370, and Raystorm Acetal were all close enough to call a tie thermally, so it’s a good baseline to run with.

  4. Lutfi says:

    erm, liquid metal is one tuf nut – it actually destroys the metal its in contact with.

    [img]http://www.frostytech.com/ArticleImages/200511/liquidmetal-destroy.jpg[/img]

  5. owcraftsman says:

    I would not say this is a huge problem but it is indicative of what we have all come to expect in terms of consistent quality for TIM. Anyone who uses TIM often has noticed how inconsistent they can be not only in terms of performance but as you allude to here density and flow. We expect that from different brands but when it’s coming from the same brand and so called formula it’s particularly disconcerting. Unfortunately I see no way to get consistent results when shelf life and consistent formulas play a role. The good news is that this is no way reflects badly on the reviewer who has taken every precaution to insure consistent results. I completely understand the need to have bulk quantities of TIM that is paid for vs supplied and that resources are limited. This is what makes your decision to use MX2 a good one. It’s not that it’s the best TIM that matters here. It’s one round up and one batch of TIM that makes the most sense and the caveat being YMMV with a different TIM. Even if you run a roundup of TIM that would only prove what that round of batches reveals and the inconsistencies will remain ever changing. That said I would love to see these TIM roundups more often as it may just divulge some consistent irregularities that point to certain manufacturers as being better that others. In the end challenging them is the only way we will see improvement in quality. One or more of the manufactures will find it in there best interest to step up and make it better. One thing for sure, the manufacturers of watercooling products are constantly improving there products to be the top dog and appear so in these reviews and there are none better, more respected or prestigious than what’s found here at Martins or Skinnee Labs. Thank You for all the hard work!

  6. Joe says:

    BTW Indigo Xtreme is no longer going to be sold under that name in the US, Maingear has taken over the OEM and it is now titled Epic T1000 from them. The pricing is the same. You can still find Indigo at Frozen and Performance until supplies run out.

  7. PunkNugget says:

    To Joe:

    Too bad, As that is is a great product. I not only know one of the owners (Chris Macris), but I was a reseller for their product line. I have to still say that it’s BY FAR THE BEST THAT THERE IS ON THE MARKET !!! NOTHING COMES CLOSE !!! Also, when I had to take off the ETI (i.e.; “TIM”), all I had to do was peel it off and that was it. No residue – NOTHING. Once in a blue moon I had to just use a my microfiber towel or the edge of my fingernail to flake off a speck here or there, but the amazing thing is I didn’t have to use any solvent AT ALL if I chose not to. My nickel plated head from my Koollance CPU-360 & CPU-370 was just as “mirror finished” as it was the day I bought it. Here’s a web link so you can see it for yourself:

    http://lab501.ro/racire/cele-mai-bune-paste-termice-metal-vs-carbon-vs-oxizi/8

    Also, the video review that you saw (above); a 4 to 6C temp drop, on your average TIM (like Arctic Silver 5) that’s just amazing… Also, not having to change it but every 2 years is also incredible. Anyone that can get their hands on this stuff now would be wise to do it. I only say this because maybe a year or two from now Maingear may stop selling it because the process of installing it alone might discourage potential buyers; watch this tutorial:

    Again, it might be a little painstaking, but I’m a complete n00b, and if I can do it… anyone can… with an hour on their hands along with some time and patience. Also, follow the vid instructions. Here’s their website and info on the product if anyone is interested – BY THE WAY, THEY SUPPLY TWO (2) of the the ETI’s (just in case you mess up on the first try):

    http://www.enerdynesolutions.com/prod_indigo.html

    And where you can go to purchase the Indigo Xtreme:

    http://www.sidewindercomputers.com/inxtforincoi1.html

    http://www.frozencpu.com/products/9686/thr-71/Indigo_Xtreme_Engineered_Thermal_Interface_ETI_Kit_for_Core_i7_LGA_1366_-_2_Pack.html

    and the “new” EPIC T1000 by Maingear (formally Enderdyne Solutions):

    http://www.amazon.com/MAINGEAR-T1000-Engineered-Thermal-Interface/dp/B006ZOPKZK

    Thanks Joe for the input and again, it’s sad that this ETI didn’t do very well in sales for the consumer as it’s done for the bigger corporations that use this ETI for multiple servers… Later… 🙂

  8. Phos says:

    It isn’t practical to use Indego Extreme in block testing, and it takes the quality of the base too much out of the equation. Plus, installing it in a multiblock loop means your graphics card or whatever else is going to heat up as well, which can easily cook a normal paste.

  9. Martinm210 says:

    Yeah, as much as I know IX is a good performer, it’s just not practical at all for testing. At $10 per mount, it would cost more in IX than some blocks are worth new. I’ve had a few blocks doing nozzle testing with 20 to 30 mounts on just one block. Granted I probably would have the luxury of doing fewer mounts since it would take most of the mounting variable out of it. But…it also isn’t representative of performance with a paste where bowing and mechanical contact is extremely important.

    No doubt IX is a great performer, but just not a good solution for testing due to high cost and non typical base performance.

  10. Derius says:

    i heard these guys are actually better and cheaper
    http://www.frozencpu.com/products/10740/thr-77/Coollaboratory_Liquid_Ultra_100_Metal_Thermal_Interface_Material.html?tl=g8c127s451
    i forgot were exactly i saw the benchmark but i rember that this beat the indigo by a degree or 2 but its kind of hard to apply