Should I change out the thermal paste, or the fan?

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510
Hello, I have this cooling dilemma I need to assess with my Core i3-4370 CPU. I apologize if I have mixed anything up.

Since I got the Cooler Master GeminII S524 Ver.2 cooler, I noticed that the idle temperatures were significantly warmer than my stock cooler: it was as low as 30 - 35 degrees Celsius (stock cooler could pull around 28 - 32 degrees Celsius minimum). My big problem is, is that the core temperatures (including the package temperatures) were just as high as how well my stock cooler could do it, which is over 60 degrees Celsius, and peaks at 65. I expect to have at least a ten degree difference (the temperature range I'm aiming for is 50 - 55 degrees Celsius max); the processor's TCase is 66.4 degrees Celsius. Since it has the Cooler Master Silencio fan, which is less noisier under load, but at the expense of less airflow, to which should I purchase: a high static pressure fan, or a fan optimized for more airflow? I am thinking of getting one from Corsair with LED lights on it (or I can get one without the LED lights).

Also, another potential contributor is likely the thermal compound, as I was forced to use the one that came with the cooler, since my brother's thermal compound was not readily available at the time. I do not want to purchase the Arctic Silver 5 and find out the thermal compound is not the problem, and I do not want to purchase the fan, yet, and find out airflow isn't the issue, either. So, please tell me: is it the fan that's the problem, or is the thermal compound the problem? Or both? If neither, then I'd have to return it for a refund and get a better one, but I don't want the Hyper 212 EVO, since it's too high profile and doesn't work very well with my RAM sticks since they have high profile heatsinks.
 
Solution
You should be pulling lower temps than a stock fan. But if you're getting a max temp of 65c at 100% usage you're doing better than most. Keep in mind, tCase and tJunction are very different numbers and you are measuring tJunction with whatever software your running and that temp maxes out at 100c.

Before you buy any new cooler, always try reapplying thermal compound and reseating the heatsink. It can make a big difference. Since you're doing testing on trying to get better temps, I'd recommend avoiding AS5. That compound has a cure time of something like 100 hours, so you wont get accurate readings right off the bad. I personally like Noctua NT-H1 for a good quality compound. As for applying it and whatnot, read through these articles...
Review of the cooler here.
According to the review your temps are a little high but not a lot.
Reapply thermal paste and reset the cooler. The direct contact heatpipes have grooves between each pipe so a little more paste is needed for full coverage. Most coolers with flat bottoms need a grain of rice size. Because of the groves you need to apply a small pea or lentil size for first application.
When reapplying the groves should already be filled so use a grain of rice size.
http://www.tweaktown.com/reviews/7342/cooler-master-geminii-s524-version-2-cpu-review/index6.html
 
You should be pulling lower temps than a stock fan. But if you're getting a max temp of 65c at 100% usage you're doing better than most. Keep in mind, tCase and tJunction are very different numbers and you are measuring tJunction with whatever software your running and that temp maxes out at 100c.

Before you buy any new cooler, always try reapplying thermal compound and reseating the heatsink. It can make a big difference. Since you're doing testing on trying to get better temps, I'd recommend avoiding AS5. That compound has a cure time of something like 100 hours, so you wont get accurate readings right off the bad. I personally like Noctua NT-H1 for a good quality compound. As for applying it and whatnot, read through these articles for a good idea of what to do and to look for in a paste.
http://www.tomshardware.com/reviews/thermal-paste-heat-sink-heat-spreader,3600.html
http://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616.html

As for buying a new cooler that will work well with high profile RAM, look at the Cryorig line up. They're a bit new on the scene but they're performance as been highly rated and most models don't overlap the RAM slots. M9i is comparable to the Hyper 212 Evo, the H7 is a step up from there and the H5 Universal is a further step up. I prefer the mounting of the M9i and the H5 to the H7 but the H7 isnt bad by any means.

Also for more than you probably ever wanted to know about the temp of your CPU, read through this guide.
http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html
 
Solution

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510


If you're really thinking of the direct contact heatpipes being separate from each other, I have used the pea method (about the size of an actual pea) on the Hyper 212 EVO for an AMD FX 8350 and I've seen better results. (The direct contact heat pipes are flushed against each other, in other words.) I might have to take another look at it, again, to clarify, though I used the Arctic thermal paste. (I think they're the same people that do Arctic Silver, but I think it's like Arctic MX-2 or something.) The intel processor's heat spreader is rather small to begin with, so I seriously doubt any more than a pea sized amount of paste is efficient enough.
 

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510


Thanks for the insightful reply in regards to sending me a few articles to look at. I have, so far, read the first part of the two, and they were very interesting. The information about intel processors have helped me learn more about specific temps about the processor, though I likely have skimmed a handful of sections of the article. The big problem I have, right now, is that it is rather time-consuming and potentially tedious to reseat the HSF due to it requiring me having to take out the motherboard, again, and having to do the same process again (which can take well over 15 - 30 minutes as it is by total). So with this information I have learned, so far, is that it doesn't exactly matter what thermal paste you use, regardless if it comes with it or not, at least aside the AS5. It's odd of you to say to not recommend the Arctic Silver 5, despite it being highly revered by a lot of customers who have bought and used this thermal compound. I have also looked at Cryorig's stuff and it looks pretty cool from a first glance or two. I can see the H7 has a rather large fan, but would that still interfere with my RAM slots? And my case claims to be compatible with coolers no taller than 170mm, at least, though I can see that even a Hyper 212 EVO would be just as much of a problem with my case as it would to my brother's. I also have to add that my motherboard is a mATX form factor (saying it for the sake of notification), so it may or may not affect how I will be able to position my cooler without interfering with my PCIe 16x slot. Correct me if I'm wrong. (Though a 140mm fan on a cooler sounds pretty cool, if you ask me. ;) )

Restating about the issue with doing the reseating the cooler and reapplying the thermal paste, I could *try* and do that, though I have this hunch that all that work wouldn't be worth it. (Meaning even with "reseating" the cooler, though I doubt this is even necessary and reapplying the paste, it may still bring about the same results.)

So in short, it's a little high, but normal when under load? (60 degrees Celsius.) It does ah heck me when it idles at 35 degrees Celsius like how my Core 2 Duo E7600 would, if not slightly less than.
 
A picture of the bottom of the heatsink.
http://images17.newegg.com/is/image/newegg/35-103-204-Z07?$S640$

The Heatpipes are against each other, but they still have small groves between each pipe.
Some are smaller than others, which is a common occurrence when the heat pipes are compressed.
That is why I stated a pea/lentil size for first applications, to fill the groves.
And rice size for resetting it as the groves will already be filled.
This is my experience over the years, a rice size application on the initial seating, will not spread out evenly to cover all of the heat pipes.
 

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510


The original 212 was really bad with groves. it had an aluminum spacer between each heat pipe.
The 212 EVO, that I have now is like your cooler, with the heat pipes compressed next to each other. But it stills has groves , but smaller.

To me, I saw little difference between the 212 EVO's and the GeminII S524 Ver.2's flushed copper heat pipes. Another thing I have to keep note is that I really don't like seeing thermal paste being squished out to the point where it'll be shown on the edges of the processor. I just do the pea method because it's more simplistic than the grain or lentil method. I also observed online that it doesn't make much of a difference, anyways, and is really user preference.

And yes, on the Hyper 212 Plus (if that's what you're referring to) I have to agree on the whole spaced out copper heatpipe problem: I found that to be rather unattractive.

Also:

That is why I stated a pea/lentil size for first applications, to fill the groves.
And rice size for resetting it as the groves will already be filled.

I have this need to clean out both the heatsink and the processor every time I either change the heatsink, processor, or thermal compound, so if I did that after the next application, that would cancel out what you basically said (or that cleaning it will have the "fill the grooves" thing to be nil).
 
Since this is a new install, you really do not have to clean between the pipes if you are using the same thermal paste.
If it was a longer install I would clean between the pipes to make sure no hardened material somehow got between the pipes and heat spreader.

I understand the need for cleaning everything out, my son has the same qualities.
Another option is to fill the groves before installation using a credit card or plastic spatula to remove any excess paste and force it into the grooves.
An older article about it here.
http://archive.benchmarkreviews.com/index.php?option=com_content&task=view&id=170&Itemid=38&limitstart=5

The Gemini cooler is better than the stock cooler temp wise, but it is not a high performance cooler. Just a lot quieter than the stock cooler.
The fan moves about 67cfm which is a lot more than the stock cooler.
So if you start looking for a replacement fan, make sure the new one moves more air or temps will go up.
 

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510


And neither the sink, I suppose? Since it's all just a hunk of metal. The main things I know about a CPU are what's the name of it, what architecture it's based off of, what is the pin number (like, in this case, LGA1150), how many gigahertz does it run, how many cores and threads does it have, and what are its features. I have never paid such attention to such details i.e. the manufacturing process or one small thing about it that may be a potential problem as you have mentioned.

Since this is a new install, you really do not have to clean between the pipes if you are using the same thermal paste.
If it was a longer install I would clean between the pipes to make sure no hardened material somehow got between the pipes and heat spreader.

I understand the need for cleaning everything out, my son has the same qualities.
Another option is to fill the groves before installation using a credit card or plastic spatula to remove any excess paste and force it into the grooves.
An older article about it here.
http://archive.benchmarkreviews.com/index.php?option=co...

The Gemini cooler is better than the stock cooler temp wise, but it is not a high performance cooler. Just a lot quieter than the stock cooler.
The fan moves about 67cfm which is a lot more than the stock cooler.
So if you start looking for a replacement fan, make sure the new one moves more air or temps will go up.

Like I have aforementioned, it is pretty quiet, and I like that. However, you stating this leads back to my question: in the event the fan is inefficient (which, by god, it shouldn't be inefficient), would a fan designed for airflow or with high static pressure be sufficient for proper CPU cooling?

I have also looked at the article you showed me, and according to this, you're saying that something like this would absolutely benefit my cooling issue? I can always take a good look at how the compound looks on my processor, and yes, this is a fresh install as of two nights ago, so I can safely assume that the thermal compound is easy to clean off.
 

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510
I just took out the cooler and saw the thermal paste was applied just fine. I came in working on the computer, knowing doing that was just a waste of time, and my feelings were confirmed by reality. I am highly apprehensive on "changing out the thermal compound" but seeing as how I fear potential inoperability (even though it's a fresh application), I might as well should clean it out, anyways.
 


Arctic and Arctic Silver are two different companies. It's not that I don't recommend AS5 in general, as it is a good basic compound, its that I don't recommend it in your case because you're trying to get instant results and AS5's cure time is something like 100 operational hours.

As for the Cryorig H7 vs Evo, the Evo will overhang the first RAM slot, the H7 will not. The H7 coolers better is is quieter than the Evo. It's also shorter than the Evo by 10mm (Evo is 155m, H7 is 145mm) and doesn't require the fan to be adjusted for RAM heat sink heights. Those are the main reasons I recommend it over the Evo. However since your case can handle 170mm height, either will work without issue. Its odd that a case that of your size doesn't have a cutout underneath the motherboard for CPU cooler install/removal. What case are you using?

As for mATX form factor making you adjust your cooler because of the PCIe slot, it depends entirely upon the cooler. The Evo and the H7 wouldn't require changing position. If you were to go with a massive beast like the Noctua NH-D14/D15 then you may have to but thats a different matter entirely.
 

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510


I have already observed the hanging over the RAM slot part from the EVO, for sure.

My case does have a cutout, actually, it's just absolutely impractical to swap heatsinks or just remove and reapply heatsinks like the GeminII while the computer is on its vertically standing position, or any position at all. Because I have to unscrew the bolts on the motherboard's backplate side, as I can't unscrew the screws from the heat sink (if I tried from the top, the bolts would just fall out, anyways, so it obviously requires me to have the motherboard out, anyways). It's also too much work to flop around the case (Thermaltake Core V41, here) so many times just to get the cooler off and on the thing, plus it's more smoother to deal with just having the motherboard out of the case and applying the heatsink on a horizontal position to begin with. I am tempted to try out the Cryorig, for sure, though I am a little pressed for money, at the moment. If this cooler does not improve after following the suggestion to reapply the thermal paste my thermal problem any further, then I may try out the Cryorig cooler you suggested (seriously, a 140mm fan sounds really cool). However, I am also feeling as if I should definitely try out swapping out the fan with a Corsair one, and see if that fixes anything. I also have to keep my GTX 960 in mind, too.

Speaking of which: the width of this cooler is (almost) a tight fit in between my card's PCB (and traces) and my what was then my rear exhaust fan; there is very little space in between the card and the cooler's heat sink where you can mount a 140mm fan on the heatsink). It also covers my rear SYS_FAN header on the motherboard, so there is no way would I be able to plug that in there, and had resulted me in removing that from my case, outright, until another cooler takes over. As for anything bigger: nah, not necessary, in my case. *cue badum tiss*

If anyone fancies getting a bit nosy about my computer, I have pictures I can post to show the new setup I have recently put together, for the sake of better education in the matter.

Update: I was actually mistaking the H7 for the H5A.
 

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510
Update: After a run on Battlefield 3, I have noticed that the package temperatures are no different than the first time I installed the heatsink. I have a feeling this heatsink's fan is unoptimized for efficient cooling. I have tried to make sure which type of fan would be highly efficient for the heatsink, and I pray that a high static pressure fan would hopefully make it better, but conversely, I would be spending more time and money on trying to swap fans on this thing than even buying a new CPU cooler, but god forbid would I ever want to stick back with the stock cooler, again, so I'd probably have to see if I can use Newegg's 30-day return policy within a timely manner (for instance, I'd probably could buy the Cryorig azaran suggested, and begin the RMA process at least after I get the cooler). Keep in mind that I had to use the same thermal compound.

This is quite a dilemma.
 

ddebroux93

Commendable
Mar 9, 2016
16
0
1,510


I just think the one I have in here sucks, at this point. :/

The fan on the GeminII doesn't help much, either.

I would be sure if this "hunk of metal" would be "faulty," there would obviously be something seriously wrong with it. Either way, I'm considering on getting a refund on this one. Will have to think this through.
 


Check your bios and motherboard settings.
My last couple boards had settings for fan control.
Example:
Low noise, setting the fan profile for quiet operation.
Normal , setting the fan for normal operation.
Performance, setting the fan for higher speeds and better cooling.
Also on some motherboards you need to "calibrate" for lack of a better description, 4 pin fans. That way the bios knows what the upper and lower RPMs the fan is capable of.

In one of your later posts you mentioned you removed the rear exhaust fan.
Do you have any other case fans? A lack or case air flow would also raise temperatures.