Microsoft and Nintendo both get it manufactured for them by IBM. Not heard that TSMC are capable of EDRAM, any links?
"Microsoft's selection of the TSMC 90nm eDRAM process for the
graphics-rich Xbox 360 is an important validation of the capability and
maturity of the technology," said John Wei, senior director of Platform
Marketing, Advanced Technology Division of TSMC. "Furthermore, the
production marks the successful continuation to an ongoing collaboration
between Microsoft and TSMC."
I think the future of AMD will be its cat line. If we see any large cores introduced in the future, they'll most likely be based off the small cat cores; from my understanding they're very efficient and well utilized, just largely execution limited. Maybe AMD will name them after big cats?
A 1.2GHz broadwell core should outperform a 2GHz cat core (A6-5200), while using less die area and less power.
And the price of that Broadwell chip would match that of the AMD cat chip?
Also, when you say it uses less die area, I guess you're assuming they are on different process nodes, right?
Broadwell ULT est. die area - 114 mm^2They are a dead end. A 1.2GHz broadwell core should outperform a 2GHz cat core (A6-5200), while using less die area and less power.
Broadwell ULT est. die area - 114 mm^2
Kabini die area - 107mm^2
Cinebench 11.5 score
Athlon 5350 @ 2.05 GHZ - 1.99
Haswell Quad @ 1GHZ - 2.01
Don't forget Airmont.And the price of that Broadwell chip would match that of the AMD cat chip?
So how come talking about Intel and Microsoft here ,on this thread ,for pages,with no relation to the Carrizo whatsoever is not thread crapping Mr Stahlhart ?!
Have you read this thread.
Or only newcomers to the forum are to be warned.
Feel free to ban me ,i couldn t care less.
Don't forget Airmont.
Broadwell ULT est. die area - 114 mm^2
Kabini die area - 107mm^2
Cinebench 11.5 score
Athlon 5350 @ 2.05 GHZ - 1.99
Haswell Quad @ 1GHZ - 2.01
Your BDW die estimate looks wrong. Die estimate is 80-85 mm² for Broadwell-Y/ULT.
Airmont will have 2.7GHz turbo clock speed, about 10% higher. But it will consume a lot less power and it will have a huge increase in GPU power.And anyway, it's uncertain how good Airmont will perform. So far the low-priced notebook and mobile CPUs from Intel have failed to impress, and have not sold in much volume. That is despite Intel year after year claiming that next year they will succeed.
Are we talking about total die size (all CPU cores + iGPU)? And has it been compensated for that they are on different process nodes? Otherwise I guess the comparison will be skewed.
Are we talking about total die size (all CPU cores + iGPU)? And has it been compensated for that they are on different process nodes? Otherwise I guess the comparison will be skewed.
It's probably wrong, I used the 2C GT3 die area. The estimate is based on this article, you also did the math in the commentsYour BDW die estimate looks wrong. Die estimate is 80-85 mm² for Broadwell-Y/ULT.
It's probably wrong, I used the 2C GT3 die area. The estimate is based on this article, you also did the math in the comments
What is wrong? We don't have 2+3 Broadwell picture, I doubt Broadwell-Y will get a 2+3 configuration.
I'm sure they'll make a 2+3 configuration just for Apple, even if no-one else gets it.
Now I'm confused: why wouldn't they make a 2+3 15W part?As a 25W or 28W TDP part called Iris but I doub't it for a 5W-15W part.
As a 25W or 28W TDP part called Iris but I doub't it for a 5W-15W part.