News Focus
News Focus
Followers 29
Posts 25865
Boards Moderated 0
Alias Born 09/11/2002

Re: Elmer Phud post# 44619

Monday, 06/25/2007 1:12:45 PM

Monday, June 25, 2007 1:12:45 PM

Post# of 152288
Elmer, it's funny you should mention this.

Hans DeVries, the very same architect who is currently in a mad rage on the Ace's forum, has done some reasonably good work in charting out die size of past and current designs.

http://www.chip-architect.com/news/2007_02_19_Various_Images.html

You can see Barcelona at 283mm^2, for example, while Penryn is at 107mm^2 (quad core Yorkfield will obviously use 2 of these and be far lower cost and higher yielding). Quad core Phenom (codenamed Agena) will be using the Barcelona die, but with Hypertransport links disabled. So it's essentially the same die size.

In terms of a dual core Phenom (codenamed Kuma), I would use the die photo as a guide. If you are keeping the L3 and the external interfaces, and only subtracting 2 cores and their L2 caches, I think a rough estimate will be in the range of 2/3 the size of Barcelona, or roughly about ~190mm^2.

AMD will also have a smaller dual core die (codenamed Rana) with no L3. Subtracting this, you're left with a little less than half of the quad core die. Let's call it 45%. That would make a die size of about ~125mm^2.

The single core (codename Spica) part is said to be based on Lima, which is the still K8 based, not K10. I imagine this would be closer to ~80mm^2, but that's just a WAG.

I think these are good enough guesses for now. To summarize (again, these are approximate guesses):

Phenom X4 nn-7xxx Series (Agena) - 283mm^2
Phenom X2 nn-6xxx Series (Kuma) - ~190mm^2
Athlon X2 nn-2xxx Series (Rana) - ~125mm^2
Sempron nn-1xxx Series (Spica) - ~80mm^2
Volume:
Day Range:
Bid:
Ask:
Last Trade Time:
Total Trades:
  • 1D
  • 1M
  • 3M
  • 6M
  • 1Y
  • 5Y
Recent INTC News