Ivy Bridge Cooling - Taming the New Heat Issues

[BackDraft]

This really isn't meant to be a lengthy explanation of the technical aspect of Ivy's thermal differences, just findings as I benchmark previously reviewed coolers on the i5 3570K. Ivy has a specific thermal footprint with the new process, and also happens to be very voltage and temperature sensitive. So, if you are looking to cool Ivy with one of the coolers I have previously benched on AMD and/or LGA1156, here are some quick and dirty findings, with max temp at max OC under load listed:

Cooler Master V6 GT - Performed as expected with excellent upper midrange performance. The V6 GT fared better on AMD than 1156, and this seems to be the case again with better Bulldozer than Ivy performance. Again, performance is very good on Ivy, it just happens to be exceptional on AMD rigs. Max Temp 67

Xigmatek Aegir - Someone at Xigmatek must have seen Ivy coming...the stacked heatpipes in the center of the contact block are right over Ivy's highly concentrated heat. Exceptional Ivy performance. Max Temp 65

Zalman CNPS11X Extreme - Very good upper midrange performance matched with great acoustics. You give up a few degrees of cooling to lose a ton of noise. Mind you, temps stayed very good even at max OC, never topping 68 degrees. Max Temp 68

Antec Kuhler 920 - Liquid with a tiny, concentrated hotspot is often a very bad mix......the Antec actually did very well on the cooling end, provided you don't mind spending twice as much for the same cooling, and dealing with more noise than you will want to stay in the room with. Performance at normal fan settings were not so desirable.... Max Temp 65 At Earsplitting Volume, 74 At Tolerable Noise Levels

Noctua NH-C14 - If you need lower profile, it will keep Ivy cool enough. But at ~$85, it is for low profile necessities only. Max Temp 70

Phanteks PH-TC14PE - Still the king by a wide margin, and showed no ill effects of the changed thermal structure. Max Temp 59

Cooler Master TPC 812 - Price/size/performance on this thing is otherwordly. It blows away every other single tower I have come into contact with...for $69 retail and only a single fan that is astonishingly quiet. Cooler Master kicked the entire cooling market square in the nuts with the TPC 812. Max Temp 63

Phanteks PH-TC14CS - Like the NH-C14, if you need low profile (or just prefer the look in the side window, see the Phantom 410 SE) it does a great job, and at $75 with much better looks than the NH-C14..... Max Temp 69

NZXT Havik 140 - Eh, still big and underperforming for its price. Max Temp 72

Scythe Mugen 3 (1 1600rpm Slipstream) - I had forgotten how good this cooler is for $49. I had also forgotten that Scythe has mastered counter-intuitive installation with many tiny pieces. Max Temp 69

More numbers will come as I go......

[BackDraft]

for which CPU are these temps for? 3570k I assume? Do you have a noctua NH-d14 for reference? because that Phantek PH-TC14PE looks like its giving noctua a run for its money. (my assumption D14 > C14) as the C14 is for having a low profile.

[BackDraft]

for which CPU are these temps for? 3570k I assume? Do you have a noctua NH-d14 for reference? because that Phantek PH-TC14PE looks like its giving noctua a run for its money. (my assumption D14 > C14) as the C14 is for having a low profile.

Yes, the 3570...thanks for pointing that out, I thought I had put that in. Fixed it.

As far as the whole Noctua/Phanteks thing....I think I read somewhere that those Phanteks are pretty good, maybe even slightly better than the Noctua NH-D14 (which, yes, is a better cooling performer than the NH-C14). Now where did I see that review.....Oh, yes, here:
http://hitechlegion.com/reviews/cooling/heatsinksfans/13963-phanteks-ph-tc14pe-cpu-cooler-review

Actually, the Phanteks dethroned the NH-D14 across the board in reviews (except one), but only by a slim margin. But, just as important in the Noctua/Phanteks showdown; Phanteks doesn't use brown fans.

The Phanteks results on Ivy were insane. But, I literally did these benchmarks all back to back in the same build in one night, in the same room with the temp the same in the room. Looking at the other numbers, my expectation on a Core i7 870 would have been around 60 degrees for the Phanteks, not 57 degrees. I'm going to rerun it to see if there was a temperature anomaly in the room at the time, or if it just really likes Ivy's thermal footprint like the Aegir does.