Ivy Bridge's heat problem is indeed caused by Intel's TIM choice

[Zardnok]

I had to go on a thermal paste run. I found some Antec Nano Diamond 7 at staples, it is expensive @ $15 but whatever. Dropped it in my Bulldozer and it cut my temps down 6c idle and about 8c load compared to Ceramique 2. I wonder what this stuff will do under the IHS.

I think the general consensus is that the Diamond stuff is too abrasive to use on a bare die. Your best bet is to pick up some Arctic Cooling MX-4 or Prolimatech PRO-PK1. Both are high-end TIM that consistently rate well and are only $10 a tube.

 

[SilverAlien81]

The thread I've linked below has everything you could want to know about delidding a 3770. It even has graphs and charts and diagrams!
http://forums.anandtech.com/showthread.php?t=2261855

 

[BonzaiDuck]

The thread I've linked below has everything you could want to know about delidding a 3770. It even has graphs and charts and diagrams!
http://forums.anandtech.com/showthread.php?t=2261855

. . . . which addresses the abrasive properties of the micronized diamond, from which you can choose yourself based on your own perceptions of risk. You may find assertions by Innovative Cooling that they've applied their formulation to similar die situations and notice no damaging effect. That also leaves the liquid metal or metal-pad solutions, which come closer to the fluxless solder Intel used for prior generation CPUs.

 

[moonbogg]

Borked by design. Its just too bad.

 

[BonzaiDuck]

Borked by design. Its just too bad.

Well, looking back, I never found it worthwhile or immediately beneficial to upgrade from one generation to a socket-compatible very-next generation. Or -- I just went ahead and built a newer machine, as I did when I built a Wolfdale following my Kentsfield system.

If Intel didn't plan its product development and commercial releases, we wouldn't now know about something called "Haswell."

There is one major competitor -- AMD. There are also other competing processor lines, but likely for more limited markets. So it's a free market, and any "barriers-to-entry" other than patents making it less so are expected. Instead, a coordination of technologies insures that the original creator of the technology -- tracing the history back as far as the 8086 -- would be the "dominant firm." But for a mass segment of the consumer/user buyer market, it seems like a duopoly situation with AMD.

I think I'm repeating others here when I say that a vast segment of our own consumer-market -- ignoring that we're "enthusiasts" -- don't care. New computers based on either Intel or AMD will roll off assembly lines, and people unfamiliar with a simple screwdriver will buy them with hardly a thought to anything other than replacing an older system or bad thoughts about an older system per problems caused by something other than hardware configuration. [In other words, the user segment not likely to know about causative distinctions.]

Those folks might actually conclude that Ivy Bridge is a significant improvement over everything; they probably don't even know the distinctions between Sandy and Ivy. The remainder of us should also see the advantages of Ivy Bridge, while recognizing the extra trouble and risk associated with reaping the advantages we might get with a cooled-down version of the die-shrink.

It's good enough that the monthly power-bill shows any relief, or that even the unmodified Ivy Bridge can "sometimes" clock as well as the Sandy with good cooling. Just not good enough if you already have a Sandy in your box . . .