Planning to OC i5-3570K, need a $30-$70 cooler

[sherlock]

Hi, I am planning on building a desktop for Gaming, here is the basic specs:

Processor Intel Core i5-3570K 3.4GHz $229.99
Motherboard Asus P8Z77-V LK $144.99
Memory Corsair Vengeance 8GB (2 x 4GB) DDR3-1600 $53.99
Graphics EVGA GeForce GTX 670 FTW $407.99
Storage Corsair Force Series 3 240GB $179.99
LITE-ON DVD Burner - Bulk Black SATA Model $17.99
Enclosure NTZX Phantom 410 $100.00
Power supply Corsair HX650W $114.99
CPUFan COOLER MASTER Hyper 212 EVO $29.99
Total: $1279.92

I plan doing only minor OCing (say 3.7 GHz) at the start but eventually I might want to OC moderately(say up to 4-4.2 GHZ) and I am planning to get a Cooler in the $30-$70 range that can keep my Ivy Bridge CPU at a safe temperature during OCing. My first choice was Hyper 212 EVO b/c the price but I also looked up Thermalright True Spirit(140mm- $40), Thermaltake's Frio($60) and Corsair H60($65).

p.s If you have experience playing BF3 while OCing please let me know what setting you are on, how much did you OC and how much did it help your FPS, I am going to be playing BF3 a lot on this rig so I could use that information.

My question is for what I need(moderate OCing) which fan is the best and which one have the best value(price/performance), If I can get better cooling from a nosier cooler I am OK with it as long as it is not obnoxiously loud.

[Jon1984]

You won't notice much of a difference in BF3 by overclocking your CPU.

http://www.tomshardware.com/reviews/battlefield-3-graphics-performance,3063-13.html

There are good reviews on the Termaltake Frio cooler. You might want to check the Noctua coolers, pricy but one of the best for sure

A small list of some excellent coolers, providing very low temps on full load compared to most coolers:

Antec Kühler H2O 620
Zalman CNPS12x
Prolimatech Black Megahalems
Thermaltake Frio Advanced
Noctua NH-C14

[derFunkenstein]

Like Jon said, you're not going to get much in BF3 out of your CPU OC, but if you want to OC then you should look into the Cooler Master Hyper 212+ or Hyper 212 EVO. Great coolers for very little money. I have the 212+ and have been extremely happy with it.

[rogue426]

To the OP : your pretty much asking the same question in this other thread.

viewtopic.php?f=33&t=83042
Your going to get the same answers there as here.

[sherlock]

You won't notice much of a difference in BF3 by overclocking you CPU.

http://www.tomshardware.com/reviews/battlefield-3-graphics-performance,3063-13.html

There are good reviews on the Termaltake Frio cooler. You might want to check the Noctua coolers, pricy but one of the best for sure

A small list of some excellent coolers, providing very low temps on full load compared to most coolers:

Antec Kühler H2O 620
Zalman CNPS12x
Prolimatech Black Megahalems
Thermaltake Frio Advanced
Noctua NH-C14

I seen that review before, however the counter agrument I got from BF3 Gamer on Symthic.com is that because the 64player matches & higher level destruction in multiplayer(espically Back to Karkand Maps), CPU will have to work a lot harder than it does in SinglePlayer Campaign(Going Hunting was used to benchmark @Toms) ,here is the full quote:

Sadly "BF3 is still a GPU intensive program. Like with most games." is an all too common misperception about a lot of games. Nvidia and ATI love when you say that too. Unfortunately the truth is that Bf3 is bottled by CPU rather than GPU. The GPU mainly benefits all the graphic effects stuff. AA, ultra settings etc. However in multiplayer its all the calculations from 64 players plus the geometry of all the destructible stuff that kills your frames via cpu. Should you doubt this you can take the Sharqi TV tower test: Climb to the roof and look out at the playing field. Type render.drawfps 1 and then render.drawperfoverlayvisible 1 into console (`) Look around until you find the point at which your fps consistently drops to the lowest. You'll notice that you happen to have a lot of destructible buildings in view and that your CPU will be stressing on the perf overlay.

You can also see drastic increases/decreases by playing with the Mesh settings. This effects draw distance and mesh complexity. On Karkand maps you will see a drastic difference in fps from low to ultra. ( I said earlier to drop the Mesh to low but medium is actually better since your draw distance is maxed at medium. any further increase just effects the details and draw distance of buildings but at the cost of huge fps) As far as ghz goes, Riesig is right when he says the its not a good indicator but only when comparing different architectures (amd vs intel) When comparing within Intel chips there is a huge difference between a chip at 3ghz and one at 5. Remember that we don't care about average fps but rather minimums (how far it dips) and to test this you can go to Sharqi tower and look at the fps you get with a 3ghz vs a 5ghz. Difference is very significant.

Of interest as well is the perfoverlay graph. One yellow line and one green. Yellow is cpu and Green Gpu. Whatever line is higher is the one that's bottlenecking your system. A good rule to remember is that clock cycles do the work. So if you have a choice between a better graphics card or CPU always take the CPU. I run SLI 680's and the benefit you get from adding that card is significant but its only about 20-25% more and not 100% like logic would dictate. Now if I want to run around with everything on ultra they certainly help but then you're introducing all this extras that just confuse the scene even more which is not desirable in an multiplayer fps.

@Santewi Yeah Post is FXAA. Textures at ultra is fine but then you're losing frames for a very small benefit in pic quality over high. Only a few frames but every one counts so I would stick with high. Mesh on high/ultra is a big big no no. Huge hit in fps especially on karkand maps. No tangible benefits as draw distance for enemies is maxed at medium. The only thing that changes is details on units/buildings and building draw distance.

p.s I got quite a few recommendations for CoolerMaster Hyper Evo 212 or 212+, seem like they are adequate enough for Moderate(around 4 GHz) OC,but I will look through the other cooler you guys listed.

p.s.2 sorry about asking the same question as my main build discussion post in Systems, I will move that post toward discuss Mobo/Ram/Monitor instead.

[Jon1984]

Well, all I can say is that I have with my system (on the signature) a smooth gameplay in Ultra detail at 1080p (with 2x AA) even in 64 maps, your definition of smoothness can be different than mine of course. I don't see tangible difference in having my i2400 at stock or with more 400Mhz on it, so to make sure you don't have any regrets just buy a decent cooler and if you don't get suitable frames overclock your CPU and GPU

[Chrispy_]

I always take "Ultra" settings with a pinch of salt. If you really spend hours comparing image quality of Ultra to High, there are some subtle differences, but not enough for me to care about. Sensible tweaking means you can normally get 95% of the image quality but 150% of the framerate.

If you just drop mesh and terrain quality from Ultra to High, you're going to make a big difference to your minimum framerates on a 64p map without really noticing anything in terms of image quality.

[sherlock]

Thanks for that suggestion, I am thinking along the same line, I will go to Ultra then disable/set to middle/low all the AA/Mesh settings since I never notice these on High/Ultra anyway. As far as cooler,I recieved feed back from a user with the same Processor/Cooler setup as me(3570K & Hyper Evo212) comfortably OC to 4.2-4.4 without significant trouble/issues and his friend with the same setup go to 4.5, that's more than what I probably need so I am all set on this front. Thanks for your help.

p.s If you guys could recommend a heat paste for me that would be great: I heard Arctic Silver 5 is pretty good but also considering IC Diamond.

[travbrad]

If you just drop mesh and terrain quality from Ultra to High, you're going to make a big difference to your minimum framerates on a 64p map without really noticing anything in terms of image quality.

Yep when I first got BF3 I switched between ULTRA and HIGH on every setting and took screenshots. Even staring closely at screenshots back-to-back I was struggling to see any difference at all.

[Chrispy_]

p.s If you guys could recommend a heat paste for me that would be great: I heard Arctic Silver 5 is pretty good but also considering IC Diamond.

Use the goop that comes with the heatsink. Fancy goops will do better than cheap goops if you are a cack-handed gorilla and spooge on far too much paste, but if you use a credit card to get a very thin, even layer on the chip like you are supposed to any paste is going to be within 0.5 degrees. I think it was Frostytech that had a really in-depth article, but basically don't waste your money on the fancy stuff. It's there for the ePeen crowd to put in their signature - along with 1.5KW power supplies, amongst other things.

I like this explanation a lot. Don't skim-read it though, I know toothpaste wins the thermal-paste group test but it is not advisable to use it on your CPU.

[absurdity]

I'm pretty sure TR did a comparison of thermal paste a long while ago, and found mostly the same thing (there is a difference, but it's minimal).

[Captain Ned]

I'm pretty sure TR did a comparison of thermal paste a long while ago, and found mostly the same thing (there is a difference, but it's minimal).

Dan's Data has used Vegemite and found it to be quite satisfactory.

[Washer]

You won't notice much of a difference in BF3 by overclocking you CPU.

http://www.tomshardware.com/reviews/battlefield-3-graphics-performance,3063-13.html

This article focuses only on the single player portion of BF3. While that's obviously to make the benchmarking much easier it also makes the article far less valuable. There is considerably more demand on the CPU playing multiplayer in BF3. I don't have the time to look up extensive testing to prove this but I can point to my own experiences.

Two systems. One is Core i5 2500K (stock) + HD6870 (stock) + 8GB RAM + Asrock Z68 mobo vs Pentium G530 (stock) + HD7850 (at 1GHz core, memory stock) + 8GB + Asrock Z77 mobo. In multiplayer the 2500K system is smoother than the G530 despite the the G530 being paired with a faster graphics card. The G530 in BF3 multiplayer is fully utilized always, it is a definite bottle neck. Game still plays ok (and I only play 64p maps) but the CPU is important in BF3 MP.

[Jon1984]

Two systems. One is Core i5 2500K (stock) + HD6870 (stock) + 8GB RAM + Asrock Z68 mobo vs Pentium G530 (stock) + HD7850 (at 1GHz core, memory stock) + 8GB + Asrock Z77 mobo. In multiplayer the 2500K system is smoother than the G530 despite the the G530 being paired with a faster graphics card. The G530 in BF3 multiplayer is fully utilized always, it is a definite bottle neck. Game still plays ok (and I only play 64p maps) but the CPU is important in BF3 MP.

You are comparing a top performer quad core with a entry level dual core, its normal that the CPU is fully utilized. I was referring that overclocking the i5-3570K wouldn't give much of a performance boost.

[Washer]

You are comparing a top performer quad core with a entry level dual core, its normal that the CPU is fully utilized. I was referring that overclocking the i5-3570K wouldn't give much of a performance boost.

Except you're basing that on a worthless article because no one plays BF3's single player more than once (or at all, I haven't played it). The system load profiles of BF3's single player and multiplayer are significantly different. The chart on the top of the page you linked to shows a dual core Core i3 outpacing everything else. It wouldn't take long playing BF3 multiplayer on those configuration to realize that's laughable. In fact, I think BF3 multiplayer is CPU dependent enough to see a FPS gain through overclocking. On your own configuration if you could go beyond a 400MHz increase I bet you could sense it (and prove with an FPS counter).

[Jon1984]

You are comparing a top performer quad core with a entry level dual core, its normal that the CPU is fully utilized. I was referring that overclocking the i5-3570K wouldn't give much of a performance boost.

Except you're basing that on a worthless article because no one plays BF3's single player more than once (or at all, I haven't played it). The system load profiles of BF3's single player and multiplayer are significantly different. The chart on the top of the page you linked to shows a dual core Core i3 outpacing everything else. It wouldn't take long playing BF3 multiplayer on those configuration to realize that's laughable. In fact, I think BF3 multiplayer is CPU dependent enough to see a FPS gain through overclocking. On your own configuration if you could go beyond a 400MHz increase I bet you could sense it (and prove with an FPS counter).

Agreed, although not tangible enough to improve the gaming experience It's already fast enough that's what I was trying to say

[Airmantharp]

Go look at the [H] for MP benchmarks of BF3. Also, you need the CPU in MP- it's the single reason I went from a 3.4GHz C2Q to a 2500k.