InvestorsHub Logo
Followers 0
Posts 67
Boards Moderated 0
Alias Born 08/17/2003

Re: Hans de Vries post# 79159

Thursday, 02/08/2007 1:48:39 AM

Thursday, February 08, 2007 1:48:39 AM

Post# of 97574
We get some interesting scaling factors for Core logic and L2 sram:
 
Logic Area Scaling 130-->90 = 0.72 instead of the ideal 0.48 =(90/130)^2
Logic Area Scaling 90-->65 = 0.70 instead of the ideal 0.52 =(65/90)^2
Logic Area Scaling 65-->45 = 0.75 instead of the ideal 0.48 =(45/65)^2

SRAM Area Scaling 130-->90 = 0.67 instead of the ideal 0.48 =(90/130)^2
SRAM Area Scaling 90-->65 = 0.64 instead of the ideal 0.52 =(65/90)^2
SRAM Area Scaling 65-->45 = 0.52 instead of the ideal 0.48 =(45/65)^2

1)The logic scaling from 90->65 is exactly the same as for AMD.

2)The Sram scaling for 65->45 is either extremely good or I am
to optimistic about the 115 mm2 die size for Penryn. (They
achieved the same sram scaling in the sram test devices so it's
probably very close.

3)The Core area's of Merom and Penryn seem exact copies if you
place one over the other and make it transparent.


Regards, Hans



Volume:
Day Range:
Bid:
Ask:
Last Trade Time:
Total Trades:
  • 1D
  • 1M
  • 3M
  • 6M
  • 1Y
  • 5Y
Recent AMD News