Which Thermal Paste to Get?

[tulx]

I used the default NT-H1 for my Noctua NH-D14 cooler. A 960T unlocked to six cores levels out at 37 C under 100% load (Prime95). I suppose that is good.

 

[LOL_Wut_Axel]

Diamond IC got me about 2C better temps than AS5 under load. And they cost about the same now. To some people 2C is nothing, but to others it is a decent jump. Regardless of what reviews say, this is personal experience which trumps any review site.

Yeah, and where did I mention Arctic Silver 5? That paste is pretty much a laughing stock today. It requires a setting time before it can perform like intended, is overpriced, and underperforming. Even the TX-2 I just mentioned is consistently better.

 

[exar333]

Yeah, and where did I mention Arctic Silver 5? That paste is pretty much a laughing stock today. It requires a setting time before it can perform like intended, is overpriced, and underperforming. Even the TX-2 I just mentioned is consistently better.

You should focus on things you are more informed on. 2-3C is a BIG deal to some people, and better compound really shines in high-ambient temp applications and high-load scenarios.

Take a cooler that gives you 2-3C better cooling, add a premium fan that also provides 2-3C, and some better compound that also saves 2-3C. Thats a HUGE temp difference.

You are also making (incorrect) broad generalizations as well. Better compound on low-power CPUs will matter less vs. higher-TDP chips. For those running power-hungry CPUs, every degree cooler you can get is important.

I can spend an extra $40 on better fans/HS to get the same temp gains I get from spending an extra $5 on compound. Pretty much a no-brainer.

 

[LOL_Wut_Axel]

You should focus on things you are more informed on. 2-3C is a BIG deal to some people, and better compound really shines in high-ambient temp applications and high-load scenarios.

Take a cooler that gives you 2-3C better cooling, add a premium fan that also provides 2-3C, and some better compound that also saves 2-3C. Thats a HUGE temp difference.

You are also making (incorrect) broad generalizations as well. Better compound on low-power CPUs will matter less vs. higher-TDP chips. For those running power-hungry CPUs, every degree cooler you can get is important.

I can spend an extra $40 on better fans/HS to get the same temp gains I get from spending an extra $5 on compound. Pretty much a no-brainer.

You should focus on reading. Tuniq TX-2 gives you lower temps than AS5 AND costs less AND doesn't require a curing time.

And the difference between TX-2 and Shin Etsu or Diamond IC is, at most, 1C. Unless you're a moron or have more money than brains you'll go for the TX-2 or Arctic Cooling MX-2.

 

[Edrick]

And the difference between TX-2 and Shin Etsu or Diamond IC is, at most, 1C. Unless you're a moron or have more money than brains you'll go for the TX-2 or Arctic Cooling MX-2.

So spending $3-$5 more for Diamond IC to save 1C-2C is being a moron?

 

[KingFatty]

In almost every situation, doesn't the temperature savings convert over to a decibel/loudness savings?

So if you save 1C, you could convert that to 1dB or whatever softer fans?

Or are people running their system so on-the-edge that 1C is the difference between stability and instability? I've always run my system where I could increase the fan a bit to bring temps down more if I wanted, but I prefer the quiet running. So for me 1C better or worse just means I get a bit more or less fan noise. But when i say fan noise, I mean the soft whisperings that are barely audible when idle.

 

[exar333]

You should focus on reading. Tuniq TX-2 gives you lower temps than AS5 AND costs less AND doesn't require a curing time.

And the difference between TX-2 and Shin Etsu or Diamond IC is, at most, 1C. Unless you're a moron or have more money than brains you'll go for the TX-2 or Arctic Cooling MX-2.

I guess I can afford an extra $3 for cooler temps? 😀

 

[exar333]

So spending $3-$5 more for Diamond IC to save 1C-2C is being a moron?

He doesn't get it.

Small gains add-up. Cooler temps and a quiter setup are always better, especially when they are almost free.

 

[LOL_Wut_Axel]

So spending $3-$5 more for Diamond IC to save 1C-2C is being a moron?

You won't be saving 2C. You pulled that out of nowhere. Typical difference is 0.5C, with the biggest being 1C. That, and proper application of the paste will make more difference anyway.

 

[LOL_Wut_Axel]

He doesn't get it.

Small gains add-up. Cooler temps and a quiter setup are always better, especially when they are almost free.

No, you don't get it. You love to exaggerate numbers to suit your arguments. 0.5C isn't 2C.

 

[Edrick]

You won't be saving 2C. You pulled that out of nowhere. Typical difference is 0.5C, with the biggest being 1C. That, and proper application of the paste will make more difference anyway.

I pulled it out of nowhere? So my personal experience means nothing? So what I see with my own eyes is not as legit as some hardware review site? Whatever.

1C-2C cooler for $2-$3 is something most people would take in a second. Short money.

And yes, this is all assuming that we have proper application of the paste.

 

[Grooveriding]

I use nt-h1 and mx-4, both seem really close in performance, to the point I am talking a 1C or 2C difference which could just be a better mount. Forced to choose I would go with MX-4 because it is easier to apply and clean off. Remember to use more paste than usual on SB-E, the chip has about 50% more surface area than SB.

I would stay away from indigo xtreme as I hear it is near impossible to conpletely remove from your chip short of lapping it. I also do not think there is a 2011 application available anyways

 

[Synomenon]

Ok, so I'll forget the Indigo Xtreme. Saw some reviews that say you have to lap the CPU and heatsink sometimes just to get the Indigo Xtreme off if you ever did have to remove it.

So, now I'm looking for Shin-Etsu X23-7921-5. Can't find it anywhere though.

 

[Ajay]

Here's a good review of some of the best TIMs out: http://skinneelabs.com/2011-tim-results/2/

 

[Joseph F]

Here's a good review of some of the best TIMs out: http://skinneelabs.com/2011-tim-results/2/

Wow, that Phobya HeGrease looks awesome.
At $10 for a 3.5G tube, I'll have to try some someday.

 

[LOL_Wut_Axel]

I pulled it out of nowhere? So my personal experience means nothing? So what I see with my own eyes is not as legit as some hardware review site? Whatever.

1C-2C cooler for $2-$3 is something most people would take in a second. Short money.

And yes, this is all assuming that we have proper application of the paste.

Your personal experience means nothing because Arctic Silver 5=!Tuniq TX-2 or Arctic Cooling MX-2, both of which are better than AS5. They cost less, too.

Both of those will have a difference of around 0.5C against the top performers.

 

[billyb0b]

I've used IC7 in the past and while it's a great TIM, it leaves a stain on the heatsink and cpu

 

[deimos3428]

I'd just get the MX-4. It's good. Whether it is the best isn't particularly relevant unless you're trying for a record.

 

[VirtualLarry]

I still use AS5. Whatever. It works for me.

 

[exar333]

I pulled it out of nowhere? So my personal experience means nothing? So what I see with my own eyes is not as legit as some hardware review site? Whatever.

1C-2C cooler for $2-$3 is something most people would take in a second. Short money.

And yes, this is all assuming that we have proper application of the paste.

There are some great results with IC7, if you Google around. I agree 100%, 2-3C cooler for a couple bucks is impossible to beat in the cooling world.





http://www.hardwarecanucks.com/forum/member-reviews/47771-ic-diamond-thermal-compound-review.html

Anywhere from 2C to 6C cooler, depending on the setup and load levels. Thats a great investoment.

 

[exar333]

I've used IC7 in the past and while it's a great TIM, it leaves a stain on the heatsink and cpu

Agreed, that can happen. Doesn't bother me though.

 

[exar333]

Here's a good review of some of the best TIMs out: http://skinneelabs.com/2011-tim-results/2/

Interestingly enough, a lot of the better ones are from the same company you buy your HSF stuff from (Prolim, Noctua, Tuniq, TM, etc).

Artic Silver, as a whole, seems to have really sifted to the bottom of performance over the years.

 

[Broheim]

can't believe nobody linked this thread yet, http://forums.anandtech.com/showthread.php?t=2181357, Idontcare does a fairly extensive comparison between TIMs in it.

 

[Synomenon]

After reading IDC's thread and the Skinnee Labs reviews I went ahead and got Prolimatech PK-1.

Thanks everyone.

 

[Piano Man]

In almost every situation, doesn't the temperature savings convert over to a decibel/loudness savings?

So if you save 1C, you could convert that to 1dB or whatever softer fans?

Or are people running their system so on-the-edge that 1C is the difference between stability and instability? I've always run my system where I could increase the fan a bit to bring temps down more if I wanted, but I prefer the quiet running. So for me 1C better or worse just means I get a bit more or less fan noise. But when i say fan noise, I mean the soft whisperings that are barely audible when idle.

You and I are on the same frequency. That's why I used Indigo Xtreme for Air Cooling. Overclocking is only good if I can't hear it 🙂