The way you avoid the actual problem that's causing this issue is really pretty funny, Hei. Oooooohhhh, moar transistors!! Yeah, almost all of which went to their primitive GPU.
The efficiency gains are on par with a planar shrink, you know this.
the real problem is actually simple: It's the fin structure and reduced surface area for heat transfer to the spreader, coupled with increased wall surface area and less spacing, resulting (in all probability) in high leakage at slightly higher clocks.
This isn't just going to go away, especially when they start piling on the cores. It really isn't going to go away when they attempt a full node shrink either.
"The efficiency gains are on par with a planar shrink, you know this."
Actually after everything I wrote and you being the only AMD fan that has an ounce of tech understanding I would have thought you would have gotten it. You claim that efficiency gain is up to snuff but not anything that planer could not do. I explained that with 3 to 4 ratio of sandybridge vs ivybridge coming from the area of a wafer.
Do you actually understand what that means as far as efficiency goes. In other words the ivybridge chip on a 26% smaller die has anywhere from 7% to 15% better performance on a vast majority of test compared to the sandybridge SoC. And when compared to Intel high end E series when using a single core test Ivybridge single core scores are even higher then the E series. It isn't until the E series two extra cores kick in does the E series best the Ivybridge. As I said simply phenomenal.
Throw in the fact that Ivybridge TDP is 17 watts less when you hit those higher performance numbers and you have with 22 NM trigate one of the wonders of the world.
It will be a very sad day for ARM and the foundries when Intel focuses Trigate upon the Mobile handheld market.
"I explained that with 3 to 4 ratio of sandybridge vs ivybridge coming from the area of a wafer."
Its a full node shrink, Hei.
"blah, blah, blah.... As I said simply phenomenal."
Its a full node shrink, Hei.
"It will be a very sad day for ARM and the foundries when Intel focuses Trigate upon the Mobile handheld market."
Wooooooooooooo.... The almighty hand of Intel. Sure. Listen, there's more than one way too skin a cat, Hei. And you still have no solid numbers on yield and never probably ever will.
The problems that Intel has going forward are not easily resolvable. 22nm planar hasn't been tried yet, never will. But 20nm planar will be here and it will have it's own set of advantages. How much of an advantage is Intel's density BTW when you have a big, slow GPU and a ton of memory to shore it up on die with it?
You're splitting hairs, there's nothing perfect in the world. All AMD needs to do is refine their X86 cores and keep the chips as cool as they can. Intel can't catch AMD's GPU, and hence AMD will save a lot of die space with a much more efficient GPU design. All the while tweaking their CPU cores.
You act like Intel has this all in the bag, I beg to differ.
"Actually after everything I wrote and you being the only AMD fan that has an ounce of tech understanding I would have thought you have gotten it" Fusion...the O wise one, the "sensei" master of tech Hei has been dissapointed in his teachings. You have failed him by not agreeing with the "all knowing" of all things.
Maybe..just maybe one day we can ALL be enlightened by hei's depth of Intel's reality of perfection. Oh..how i wish I could understand hei's biased veiw of Intel, then maybe I COULD be blessed by hei as the board tech genius.