Noctua NT-H1 Thermal Paste Review

hardnrg - 2008-06-09 11:45:54 in Cooling
Category: Cooling
Reviewed by: hardnrg   
Reviewed on: July 13, 2008
Noctua
Noctua
Price: $8.99

Introduction:

Hot time, summer in the city, man my load temps sure are sh....yes, it's that time again, summer, the worst time for overclocking, unless you live in a meatlocker all year round. "Oh, I'll be safe, I don't overclock, I like my computer to be silent." Well, the cooler your CPU runs, the slower the fans need to turn, so you're not so safe after all hey Mr. Smarty Pants?

Noctua is a manufacturer that caters to both extremes, often achieving both high performance cooling and also low noise. This is thanks to its industrial ties with the heatsink and fan manufacturer, Kolink, and The Austrian Institute for Heat-Transmission and Fan Technology. The marriage of engineering and manufacture expertise, together with continuing research and scientific backgrounds, results in the progressive design of computer cooling products.

With a range of CPU heatsinks and fans, it makes sense that Noctua would develop its own brand of thermal transfer compound. Most other cooling manufacturers also carry their own brand of thermal paste, but how do they compare to each other, and how does this Noctua NT-H1 compare to thermal compounds from the companies that specialise in nothing else other than thermal paste?

 

Closer Look:

Ok, so here we have quite a large sized syringe of thermal paste presented in retail packaging ready to appear on pegs at your local computer store. As with all the Noctua products, there is an overview of the key features on the front, and more detailed specifications and explanations on the back.

 

The features that stuck out to me the most were "top-performance right from the start," meaning that there is no break-in period, "not electrically conductive," meaning there is no danger of short-circuits from the paste making contact with the motherboard, CPU, GPU, etc, and finally, the "suitable for compressor cooling," allowing you to use sub-zero cooling methods.

 

Closer Look:

If you've been building computers for a while, you probably already have your own favorite brand and type of thermal paste. But why do you like one over another? Obviously, performance is important, but I like to look at other factors when it comes to deciding on which thermal compound to use.

Probably the second most important factor to me is consistency (the texture and viscosity of the stuff). Years ago I made the mistake of buying some Coolermaster thermal paste as it was the only thermal compound available at a store that promised higher performance than generic white silicone-based compound. The Coolermaster stuff was like the dried up peanut butter you get at the side of the jar when there's barely any left, truly awful, and a complete nightmare to spread on an Athlon XP CPU die.

Even now, with pretty much everyone keeping the Core 2 heatspreader intact and the application directions for CPUs being variations on the blob-and-squeeze method, you still need to be able to spread the paste around for Northbridge and Southbridge chipset dies, GPUs, GDDR modules, MOSFET/PWMs, etc.

Here are the thermal compounds used in this review, and in the second picture is a visual comparison of what the paste is actually like. From left to right, and reading down like a book, you can see: Arctic Silver 5, Thermalright (Type 1), Thermalright (Type 2), (NOT USED: unbranded silver paste), Evercool 420, Arctic Cooling, Vantec, Arctic Silver Ceramique, and Noctua NT-H1.

 

 

After several attempts to get equal amounts of paste on the CPU, it was clear that two of the white pastes, Thermalright (Type 2) and Arctic Silver Ceramique, sank lower to the surface due to a lower viscosity, almost approaching a liquid state. Arctic Silver 5 and Noctua NT-H1 were both notably thicker, but the Arctic Silver 5 appeared wetter than the Noctua NT-H1. With Noctua's offering being the most viscous paste in this group, I wondered if it would be difficult to spread thinly on the bare dies of chipsets and unlidded CPUs.

You might be wondering about the choice of comparison thermal compounds in this review. Arctic Silver 5 is a thermal paste that I buy separately and have not seen included with any cooling products, whereas the other thermal compounds in this review are included with high performance cooling products. Thermalright used to ship the Type 1 paste with its heatsinks, and the most recent heatsink I bought, the Ultra 120, came with the Type 2 paste. The Evercool paste comes with the VC-RE chipset heatsink, Arctic Cooling paste came with one of its GPU Silencer heatsinks, Vantec's paste came with one of its Iceberq chipset heatsinks, Dangerden ship its waterblocks with Arctic Silver Ceramique, and finally the Noctua NT-H1 paste is included with the NH-U12 CPU heatsink. So, the majority of thermal compounds in this review are the same brand as the heatsinks that have thermal paste as part of the product.

While not everyone reapplies thermal paste to their chipsets and other hardware, I feel that it's an important aspect to look at for the NT-H1's evaluation.

 

Specifications:


Volume 1.4ml (for at least 15 applications)
Specific Gravity 2.49 g/cm³
Colour grey
Recommended storage time (before use) up to 2 years
Recommended usage time (on the CPU) up to 3 years
Peak operating temperature -50°C to +110°C
Recommended operating temperature -40°C to +90°C

Features:



Testing:

Testing Setup:

Comparison Thermal Compounds:

 

Overclocking:

Overclocked settings:

As well as stock voltage and speed on the CPU, I tested the highest 24/7 stable CPU overclock for a higher heat load to compare the thermal transfer performance.

 

Benchmarks:

 

Testing:

OCCT v2.0 Perestroika supports multi-core processors and will produce the most CPU heat of any stress test. Prime95 and its variants can produce the same heat load, but only after a longer period of time. OCCT has a standardised test that lasts for one hour, and produces temperature charts at the end of a successful run. The mean average of the CPU temperature over the second half of the test was recorded as a meaningful sustained heat load.

 

Settings:

 

 

 

Well, I wasn't expecting much different at stock speed and voltage, but then again, I haven't used half of these thermal pastes on a CPU before. I was surprised by the difference in load temperatures of the Noctua NT-H1 compared to the worst in the group, the Vantec thermal paste, 6°C difference just from thermal paste!

The gap widens when overclocked and overvolted, and while the NT-H1 is neck and neck with the AS5 at stock, the NT-H1 takes the lead when the CPU is producing a massive heat load. The next nearest thermal paste after AS5 is Thermalright's "Type 2," which results in a 4°C higher load temperature. The other pastes in this review are even worse, with the Vantec paste giving a whole 11°C higher load temperature!

 

Conclusion:

I wasn't expecting the Noctua NT-H1 to out-perform Arctic Silver 5 (AS5) during the tests, but it did when the system was overclocked and producing the highest heatload, albeit by a very slim margin.

The reasoning behind my doubts is the NT-H1 paste appears thicker than AS5, and I thought the resulting layer of thermal compound material between the CPU and the water block would be thicker when using Noctua's paste. This isn't the case and both NT-H1 and AS5 spread out to a very thin layer.

When using the NT-H1 on bare silicon chipset dies, I also thought the thickness would be a problem after my experience with Coolermaster's Premium Thermal Compound. The NT-H1 is quite viscous, certainly more than all the other pastes in the test, but the consistency is extremely smooth, so it is very easy to spread over a silicon die using a razor edge, quite effortless to achieve a thin and even layer in fact.

The biggest surprise during testing was just how much better the NT-H1 is compared with other cooling companies thermal compounds. Granted, some of the companies have newer, improved thermal pastes, but the large difference in performance really underlines the importance of choosing an efficient thermal paste, especially for overclocked machines.

The NT-H1 is currently $8.99 at a number of US retailers. AS5 has been around a lot longer and is available at a much larger number of retailers at price ranging from $4.99 to 9.99 for the standard size syringe. Comparing the NT-H1 to the standard size AS5 in terms of quantity, the 1.4ml of appears to be just slightly more volume than the 3.5g of AS5. It's very close, and Jakob from Noctua cleared this up for me: "AS5 comes in tubes of 3.5g. The density of AS5 is 4.1g/cm³, so it only offers a volume of about 0,85cm³ (0,85ml). NT-H1 has a density of 2.49g/cm³, so our 3.5g tube gives you 1.4cm³ (1.4ml) of paste to work with. So while NT-H1 may be slightly more expensive, you should also get more applications out of one tube." - This works out to around 65% more thermal paste in favour of the NT-H1.

So, the NT-H1 is a better performer than AS5 (only just), but you can find AS5 cheaper in the US, while in the UK, the price difference is a lot smaller. AS5 can be found as low as £4.00, and NT-H1 at £4.25. But, with the NT-H1, you get substantially more thermal compound than the AS5 3.5g tube.

As the NT-H1 beats AS5 by 1°C, and you get more paste for a similar price, the Noctua NT-H1 is the better deal for CPU cooling.

Which thermal paste do I personally prefer? It's actually still AS5, even though it loses out to NT-H1, the consistency is wetter than NT-H1. Spreading AS5 is like spreading mayonnaise, whereas NT-H1 is sort of like a very smooth clay texture. The thinner consistency of AS5 makes it easier to apply in situations that require spreading (Northbridge, Southbridge, GDDR, GPU, etc) so while NT-H1 takes the performance crown for CPU cooling, it's not quite as good as an all-round thermal compound solution in my opinion. (Update: since the production of this review sample, Noctua have modified the formula of NT-H1 slightly, so that it is easier to spread).


Pros:


Cons: