CM Hyper 212+ + i7 930?

[A_Dying_Wren]

Hi guys,

Looking to build a computer within the next week from scratch. Will be my first proper computer ever so quite excited about it. The specs I'm looking at are:

Powercolor 5870 or 5850 (probably the former which I shall hope to xfire later)
HAF 932
Windows 7 x64
6GB Corsair XMS3 CL7 1600
Corsair HX750
i7 930
Either ga-x58a-ud3r or asus p6t SE

I think this set-up is quite standard so im not too worried about it although please do point out any horrendous flaws. I'll of course need an aftermarket cooler to make it worthwhile. What I'm not sure is would the CM Hyper 212+ be sufficient to get a good 4GHz+ overclock or would I need to upgrade to megahalems to really benefit?

Thanks,
Andrew

 

[Markfw]

megahalems all the way. For 4 ghz, you really need it, unless you are very lucky on your chip, and can run a very low vcore on 4 ghz.

 

[A_Dying_Wren]

Would a Noctua D-14 also do the trick? It's tricky because I have to spend no more than 95 dollars on the whole cooling set-up if I go for CL8 ram instead (105 if I go for CL9 ram). I'm going to import my ram and cpu cooler from the US and I have to pay tax locally if it costs more than US$285

 

[nyker96]

I'd recommand zalman performa, new champ for price/performance. megahalem level performance for about 40 bucks.

 

[Markfw]

Would a Noctua D-14 also do the trick? It's tricky because I have to spend no more than 95 dollars on the whole cooling set-up if I go for CL8 ram instead (105 if I go for CL9 ram). I'm going to import my ram and cpu cooler from the US and I have to pay tax locally if it costs more than US$285

A megahalem is $60 at svc.com

http://www.svc.com/megahalems.html

 

[Deleted member 4644]

Prolimatech Megahalems (50CFM) high 16.8 10.2 45.1
Xigmatek HDT-S1283 high 16.8 10.7 50.7
Coolermaster Hyper 212 Plus high 16.9 10.9 50.1

The above data shows the CM is within .1 degree of the megahalems from the site linked above.

That wouldnt be enough difference for me.

 

[Markfw]

Prolimatech Megahalems (50CFM) high 16.8 10.2 45.1
Xigmatek HDT-S1283 high 16.8 10.7 50.7
Coolermaster Hyper 212 Plus high 16.9 10.9 50.1

The above data shows the CM is within .1 degree of the megahalems from the site linked above.

That wouldnt be enough difference for me.

I am not sure exactly what link you are talking about, but I don;t think that at 4 ghz, a hyper 212 could come close to a megahalem.

NOT STOCK

 

[OILFIELDTRASH]

Had a cm 212+ and a megahalems. There is no comparison. At all. Megahalems cools 10+C minimum.

 

[Suspicious-Teach8788]

Noctua D14 blows away Megahalems... ok not blows away but inches by..

 

[A_Dying_Wren]

Prolimatech Megahalems (50CFM) high 16.8 10.2 45.1
Xigmatek HDT-S1283 high 16.8 10.7 50.7
Coolermaster Hyper 212 Plus high 16.9 10.9 50.1

The above data shows the CM is within .1 degree of the megahalems from the site linked above.

That wouldnt be enough difference for me.

I guess these measurements would be from frostytech? The site superficially seems reliable however I can't believe the difference between coolers is that small or people wouldn't bother with anything other than the 212+.

But yea, thanks for the feedback guys. I think I'll go with the megahalems if I can find a decent and cheap fan for them. Could someone please recommend me one? I don't think the fans + $60 will push the price beyond the noctua d14.

The zalman performa is interesting but is more expensive than the 212+ for me ($34 purchased locally vs $40 + shipping from US) but seems to be in the same league as the 212+. Its a bit more convenient to get the 212+ too. Thanks for the tip anyway.

 

[Borealis7]

you cant compare frostytech to anything because they dont use a CPU for testing heat output. instead they have a plate which generated 140 W (simulate quadcore) or 85 W (simulate dualcore) of heat and then they measure.
i trust frostytech, but for real world numbers you need other sites.

 

[jvroig]

you cant compare frostytech to anything because they dont use a CPU for testing heat output. instead they have a plate which generated 140 W (simulate quadcore) or 85 W (simulate dualcore) of heat and then they measure. i trust frostytech, but for real world numbers you need other sites.

I have been made aware of this fact before, but what I don't get is why their relative results still don't jive with the real world. They use heatplates, fine. But since they use heatplates consistently for all their tests, then the performance difference between coolers that they get should more or less match up with the cooling performance difference between the same coolers in other sites that use processors.

Say, CM212+ and Megahalems. 140W heatplate vs 140W CPU (Phenom 9950?). If their heatplate tests say there is only 4C difference between the 212+ and the megahalems, then that should also more or less be the same difference in temps on the test that uses real 140W processor.

But frostytech's numbers are too close to each other. The CM212+, for example, is just 4C shy of the best cooler in their archives, and barely 2C worse than a megahalem. Yet, in real life, we have people who observe a no-contest difference between them, more like 10C or something equally significant, not just barely 2C.

Perhaps there is something more about their testing aside form heatplates that makes their results a little off?

 

[A_Dying_Wren]

Hmm... I looked around the site and i couldn't find where it stated how long the set-up was run for. It if was just for maybe half an hour that might explain the low differences.

 

[Borealis7]

perhaps running a processor OCed at full load produces more heat than the plates FrostyTech use. for instance look at this Legion Hardware test:
http://www.legionhardware.com/articles_pages/thermolab_baram_2010,5.html

assuming ambient temp of around 20C (lowest idle temp they get is 24C so it seems probable), the CM 212+ has a delta of 34C at legion hardware, while frostytech list it at 17C rise for the quadcore (150W heat output).

so unless someone doesn't have a clue at what they're doing, it must mean that the OCed i7 965 produces more than 150W of heat and thus the tests aren't comparable.

i'll fish around for some Stock speed testing.

EDIT:
and here they are:
http://www.hardwarecanucks.com/foru...rmalright-venomous-x-cpu-cooler-review-8.html

now this looks more in line with Frostytech!
about 20C rise from idle to load, putting CM hyper212 + very close to Zalman CNPS10X and the Megahalems.
so i'd say Frostytech is good for temps at stock speed (and most likely outside of case too), while OCed temps are YMMV.

and from HardwareCanucks, the Megahalems does pull away from the CM Hyper 212+ at OCed speed.
in any case, the CM 212+ looks good in every test, specially for its price.

 

[OILFIELDTRASH]

I own both the megahalems and the 212+. I don't care what frostytech or whoever says they are close it's a crock. I have no vested interest in any computer parts period unlike a tech site. I'm just a tech enthusiast. I don't care how you mount it, apply thermal paste, or put 3000rpm fans on the 212+ it doesn't come close to the megahalems. If you left them at stock speed I'm sure they are within a couple degrees. You push the i7 to 4.0ghz with 1.35v, which is probably close to what you need for a 100% stable overclock, and the temp difference will widen by a big margin. I've had two different i7 920's and now have 2 different i7 930's. Maybe I'm just unlucky but you're not gonna get 4.0ghz without 1.3v min. Sure you can boot into windows with less volts and take a screenshot but you're not gonna run prime95 24 hours. The 212 while a good cooler isn't a heatsink for the more extreme overclocks. Why would anyone buy a megahalems for $65 if we could come within a couple degrees with a $25 heatsink? It isn't happening guys.

 

[A_Dying_Wren]

OILFIELDTRASH and anyone else, I think I'll go with the megahalems but could you guys please recommend a fan to go with them? Loudness isn't too important. A good price/performance ratio would be paramount.

 

[OILFIELDTRASH]

This is a nice quiet high airflow fan
http://www.newegg.com/product/produc...82E16835103061
these are great but out of stock
http://www.newegg.com/Product/Produc...-015-_-Product

 

[n7]

High end air coolers compared in DETAIL!
http://translate.google.com/transla...prolimatech-megashadow-si-thermalright-ifx-14

 

[ZimZum]

+1 on the Noctua D14.

Megahalems are good but the added cost of a fan has to be factored into their price. Plus your going to need to 2 fans if you you want to get D14 level performance.

 

[Markfw]

+1 on the Noctua D14.

Megahalems are good but the added cost of a fan has to be factored into their price. Plus your going to need to 2 fans if you you want to get D14 level performance.

I just looked. You can get a megahalem for $60 + $5 on a fan from svc.com. The only place I find the D14 is newegg (places I trust) and its $98 shipped ! I don't think its that much better....50% more for 2c difference on a 4 ghz I7 930 ?

 

[v8envy]

svc also has the venomous-x RT for $69 -- a hair better mounting system than megahalems and includes a 120mm fan and shroud. The prolimatech armageddon *140*mm heatsink looks pretty nice for $68 as well. I might have to get one of those.

There are a few i7s which will hit 4ghz at 1.1 volts. Mine won't, even though it'll run stock all day long at .875v. One of those *might* be fine with a 212+, but your odds of scoring such a low voltage overclocker are approximately zero.

 

[ZimZum]

I just looked. You can get a megahalem for $60 + $5 on a fan from svc.com. The only place I find the D14 is newegg (places I trust) and its $98 shipped ! I don't think its that much better....50% more for 2c difference on a 4 ghz I7 930 ?

SVC has the D14 for $79. Difference of $15. Less than that if you add a 2nd fan to the Megahalem. Is the increased performance worth $10-$15? It is to some, as they pay similar premiums for Thermal Compounds that have even less of a performance delta.

 

[Suspicious-Teach8788]

 

[nyker96]

I guess these measurements would be from frostytech? The site superficially seems reliable however I can't believe the difference between coolers is that small or people wouldn't bother with anything other than the 212+.

But yea, thanks for the feedback guys. I think I'll go with the megahalems if I can find a decent and cheap fan for them. Could someone please recommend me one? I don't think the fans + $60 will push the price beyond the noctua d14.

The zalman performa is interesting but is more expensive than the 212+ for me ($34 purchased locally vs $40 + shipping from US) but seems to be in the same league as the 212+. Its a bit more convenient to get the 212+ too. Thanks for the tip anyway.

ok, first when you compare top coolers, look at reviews that use a i7 @ 3,6+ under full load. This is because under low heat situation, a medium range cooler like hyper 212+ perform on par with a cooler like zalman performa//megahalems//TRUE//VaporX//Mugen 2 etc. but what separates them is at high heat situations especially on a max overclocked i7.

if you can afford something like maga or D14 and love to use your own custom fans etc of course you cannot go wrong with either of these premium coolers. but if you are on a budget and want a universal cooler that comes with everything, one package that has all with no frills and can perform on par with a super cooler like or maga//vaporX in high heat situations like i7@4+. You better consider either zalman performa//Mugen2. they are more expensive than 212+ but that extra money is well spent. there's no chance 212+ can deal with that kind of heat. BTW you can get performa at newegg for $40 shipped.

 

[OILFIELDTRASH]

I'm calling bs on those charts. Megahalems 2c difference vs H50. I've seen some members on here say that they've used both and there was a much bigger difference than that. You've got to take these review sites with a grain of salt. Where do you think they get the money to stay in business? The only site I believe to be 100% accurate is anandtech.