SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Advanced Micro Devices - Moderated (AMD)
AMD 216.95+0.8%10:17 AM EST
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: milo_morai who wrote (75562)3/25/2002 10:15:43 PM
From: YousefRead Replies (6) of 275872
 
Milo,

Re: "If the yields are better then Intel's that's world class, no?"

Depends on the die size, Milo ... This statement is only true for AMD
die sizes that are equal or larger than INTC's. Here are the two
most commonly used equations for calculating yield from defect density
and die size -->

Murphy's DD (Mur)

Yield = [[1 - e^-AD]/AD]^2 ... A = chip area in cm^2 ... D = defect density in defects/cm^2

Bose-Einstein DD (BE)

Yield = 1/[1 + A(D/n)]^n ... A = chip area in cm^2 ... D = defect density in
defects/cm^2 ... n = number of layers = ~20 for .13um

So let's generate some examples -->

Defect Density ... Chip Area........ Predicted Yield
... .25 d/cm^2 ........ .5 cm^2 ..... 88% Mur ... 88% BE
... .25 d/cm^2 ........ 1.0 cm^2 .... 78% Mur ... 78% BE
... .25 d/cm^2 ........ 2.0 cm^2 .... 62% Mur ... 61% BE
... .25 d/cm^2 ........ 4.0 cm^2 .... 40% Mur ... 38% BE

... .5 d/cm^2 ........ .5 cm^2 ..... 78% Mur ... 78% BE
... .5 d/cm^2 ........ 1.0 cm^2 .... 62% Mur ... 61% BE
... .5 d/cm^2 ........ 2.0 cm^2 .... 40% Mur ... 38% BE
... .5 d/cm^2 ........ 4.0 cm^2 .... 19% Mur ... 15% BE

... 1.0 d/cm^2 ....... .5 cm^2 ..... 62% Mur ... 61% BE
... 1.0 d/cm^2 ....... 1.0 cm^2 .... 40% Mur ... 38% BE
... 1.0 d/cm^2 ....... 2.0 cm^2 .... 19% Mur ... 15% BE
... 1.0 d/cm^2 ....... 4.0 cm^2 ..... 6% Mur .... 3% BE

A number of interesting conclusions can be drawn from this data.

1) For the same (Area)*(defect density), the same yield is
predicted.

For example .25 d/cm^2 on 2.0cm^2 chip = .5 d/cm^2 on 1.0cm^2 chip = 62% yield

Which is better ... obviously the lower defect density and larger
die size is better than the other (higher defect density
and smaller die size).

2) For low defect density (~.25 d/cm^2) the penalty for larger die size (factor
of 2 from .5cm^2 -> 4cm^2) is not as great as for processes running at
higher defect density = 1.0 d/cm^2 (factor of 10 from .5cm^2 -> 4cm^2).

3) Defect density MATTERS !!!
(Albert ... Please pay attention <ggg>)

Make It So,
Yousef
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  

HomeMailHotSubjectMarksPeopleMarksKeepersSettings
Terms Of UseContact UsCopyright/IP PolicyPrivacy PolicyAbout UsFAQAdvertise on SI
© 2025 Knight Sac Media.  Data provided by Twelve Data, Alpha Vantage, and CityFALCON News