TIM or Thermal Interface Material, Thermal Grease, Thermal Compound, Gunk etc… It has many names, is very important and there are several types. However you don’t want to be spending £2.17 on a tube just to pay an additional £4.59 postage (Scan) or £1.21 for a tube and an additional £4.99 for postage (YoYoTech). So what can you pick up in store at PCWorld or Maplin and is it any good?
When it comes to Thermal Interface Material, TIM from now on, your options are far from limited. There are loads of products available ranging from simple Silicon based greases through to the highly developed and more expensive silver containing pastes. Amazingly enough a silver based paste can be bought these days for as little as ~ £2 online.
One thing that unites all TIMs though is the relative low cost of the product. When ordering your next tube of TIM your unlikely to pass the £50 mark that will get you free delivery from ebuyer unless you’re buying other stuff too. This means that buying a tube from an etailer will incur a postage charge and from some places that can be as much as the product itself, if not more. Although Amazon are currently offering free delivery on everything no matter how small.
So for this reason I’ve commonly bought my TIM from either PC World or Maplin. This way I can walk in off the street whenever I like and buy it there and then with no need to wait in for a delivery, no p&p charges and no driving miles to reach the ‘nearest’ CityLink depot to pick up a missed delivery!
However, buying from Maplin or PC World massively limits your buying options and I begun to think that maybe I was missing out. Unfortunately my Akasa AK-450 TIM was not in any comparison reviews either so I couldn’t know for sure. So when I ordered a motherboard from Scan I added a tube of Thermalright Chill Factor 2. Of course I didn’t just rest safe in the knowledge that I’d bought my thermal compound on line, I wanted to test and compare it to the one I already had. Anyway, so I got thinking and decided to buy a few alternatives from Maplin too just for comparison purposes.
So here’s the list of the contenders, where I got them from, how much you get and the price I paid.
|Akasa AK-450 silver based paste||PC World||£5.99||5g syringe|
|Antec Silicone Grease||Maplin||£2.99||three 1g sachets|
|Antec ‘Formula 5′ silver based paste||Maplin||£9.99||3.5g syringe|
|Oyatech silicone grease||Maplin||£1.99||0.5g syringe|
|Thermalright Chill Factor 2||Scan||£4.57 + p&p||4g syringe|
I’ve also included two pastes that were bundled with coolers for the purposes of comparison. These are the Titan Royal Grease that you get a tube of with the Titan Fenrir cooler and also the standard tube of grease that comes with the Zalman 7700 AlCU cooler, each of these has about enough for 2-3 applications.
This table will be out of date as soon as I post this article if not before. This is because stockists change their prices and their stock holdings frequently but it was correct at the time I bought them and hopefully there won’t be many differences.
The Testing Process
In order for any of this to make sense there needs to be a sensible and robust testing process. I’ve decided to use a spare PC setup as a test bench (i.e. no case) so that the CPU heatsink and fan can function to its full ability without relying on case airflow. All testing has been performed with the ambient temperature recorded such that resulting data can be shown relative to ambient rather than as absolute temperature, this effectively allows for fluctuations in ambient temperature.
All compounds are applied in the same way and temperature recorded at idle and then the maximum temperature recorded whilst at load. For this testing, load is defined as Prime 95 on both cores running the preset ‘Small FFT’ test as this provides the highest and most stable CPU temperatures.
The majority of thermal compound producers recommend applying a small blob to the centre of the CPU’s heatspreader and then spreading it out evenly and thinly. So this is the method that I’ve used for this article and I’ve kept the procedure the same throughout all the testing. There are of course many other methods for applying these compounds and different people swear by certain methods but apppliction style is not part of this article.
It’s also relevant at this point for me to state what the testing rig consists of and make comment on the CPU cooler used. It is based on an Abit F-I90HD motherboard which is a 775 socket board with an ATI Radeon Xpress 1250 chipset. In the 775 socket there is an Intel E2160 CPU overvolted to 1.55V and overclocked from the stock speed of 1.8 GHz up to 3.0 GHz. The RAM used is 2 GB of OCZ DDR2 800 MHz ATI Certified and this is kept at it’s stock speed.
The CPU cooler is a Zalman 7700AlCu, it does not have heatpipes and has a well finished base, I’ve not lapped the base to a mirror finish or anything but it was obviously well finished in production. This is significant because depending on the base finish and efficiency of the cooler the Thermal Compound will be more or less significant in the overall cooling process.
So onto the test results then. Well I must say I was quite surprised, the gap between the more expensive silver based pastes and the cheaper silicone greases was much smaller than I had expected.
Load Temperatures relative to Ambient
As you can see the Antec Formula 5 came through as the outright winner and indeed this is the highest priced product in my test. However it didn’t come through anywhere near as far ahead as the comparative prices might suggest, I’m not sure that 1 or 2 degrees of extra cooling would make a huge difference in either component life or overclocking, or at least not in my experience.
Ease of Application
If the paste and greases perform much the same with only a degree or two to separate them then we may as well give a good regard to how easy it is to apply the said paste. Well the Titan grease is not a grease at all and much more like a silver based paste in both colour and consistency whereas the Akasa AK450 is more like a grease than a paste. Apart from those two generally throughout this test I found the greases to be much easier to apply being slightly thinner in consistency. Of course even the most difficult paste still wasn’t exactly a pain to apply and if you’re only going to apply it twice a year then it’s shouldn’t be too much of an issue.
I do think though that you can get away with using a thinner layer with the more grease like compounds and as such you’ll use up your tube less quickly so it’ll be better value for money.
From these results it’s hard for me to make much in the way of comment, I think it’s better for you to read the article and come to your own conclusions really. For instance the Oyatech grease is the cheapest and scores in the middle of the pack, however you barely get any in the tube, only 0.5g in fact. If all you need is a small splodge to help out a friend or relative in an emergency then it does its job OK. If you regularly fit and re-fit heatsinks then you’ll be wanting a larger pack and to be honest, £10 isn’t a great amount of money so you may as well buy the test leader, the Antec Formula 5 if you need some paste and don’t want it mailed to you.
You will notice of course that I’ve titled this section ‘Initial Conclusion’ which kind of suggests that I have more to add. Well that’s true, in fact I have an additional cooler that I intend on using with the same pastes to see if the spread of results changes at all or if its still all much the same. That cooler is the Titan Fenrir and it has four large direct contact heatpipes for the pastes to try to marry to the CPU heatspreader.
So that’s it … for now. Expect a second page to be added to this article once I’ve completed the testing. One thing that I can reveal right now though is that on the same test rig the Titan Fenrir reduces the temperatures by more than 10°C over the Zalman 7700, I was quite surprised at the extent of the difference.