|
|
|
|
|
|
|
| Page(s): 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 |
| If you see a loading image, please wait or try again by refreshing this page after sometime. |
| Note : Best viewed in FF3 or above, IE7 or above |
| |
| Author: Benjamin Bunday (Fellow) |
Reads: 3410
Pages Viewed: 3415
|
| This paper was posted by
 |
Benjamin Bunday (Fellow)
|
|
|
|
|
Domain:
High Tech; Category:
Semiconductors; Subcategory: Reverse Engineering
Upload Date: 6th-Apr-2007 Click On Above Link To See More |
|
Members having similar interests like you
|
|
Other similar topics highly read
|
|
Recent Press Releases
|
|
| Short URL: http://electronics.wesrch.com/pdfEL1SE18ZMX163 |
|
| Comments below ordered by Most Recent First |
Benjamin Bunday
Fellow
|
Comment : Again, thanks for your comments--sorry for the multipost rant,
but it was the only way around the 500 character limit. NOTE TO
ALL READERS ON MULTI-POSTS BELOW--in reading my comments below,
read them bottom to top (chronological) order--otherwise they
look like so much drivel...
|
|
|
|
|
|
Report Abuse
|
|
|
Benjamin Bunday
Fellow
|
Comment : This is why we believe perceptions must change, and the NVA idea
changed to one that "too much metrology" is what is NVA; there is
a basic amount of metrology which is very necessary to run your
process which is of very high value. Collecting the wrong
information or too much information is what is to be avoided to
save money, but the guidance of your other process tools into
productive utilization (not producing scrap) is a fundamental
need for getting top value out of your fab as a whole.
|
|
|
|
|
|
Report Abuse
|
|
|
Benjamin Bunday
Fellow
|
Comment : Post etch CDs can also be used to tweak the implant/anneal to
fine tune to electrical targets. This is a prime example of
metrology creating value (analogous to "playing offense"). Same
with development of OPC models--another prime example of
metrology having an opportunity to play a decisive roll in the
success or non-success of a product. APC, OPC and DBM are key
avenues where metrology can do this, which are fundamentally
different from the past uses of metrology.
|
|
|
|
|
|
Report Abuse
|
|
|
Benjamin Bunday
Fellow
|
Comment : APC changes this analogy to that of a guided missile, with
mid-course corrections to always hit the target (smart bomb, the
guidance system is a metrology package); the metrology after gate
litho is used in the APC loop in the paper to tweak parameters in
the etch to achieve tighter product distributions.
|
|
|
|
|
|
Report Abuse
|
|
|
Benjamin Bunday
Fellow
|
Comment : I have a "non-politically correct", military analogy--in WW2 if
you bombed a target you took in 1000 bombers on a raid, then saw
where they went (after the fact) and evaluated if target was hit.
This is analogous to the old metrology model, executing a process
and then seeing if you are in spec or not, and rework where
necessary.
|
|
|
|
|
|
Report Abuse
|
|
|
Benjamin Bunday
Fellow
|
Comment : Also your litho example: while the litho adds the pattern and
the metrology does not, without the metrology the litho adds
"garbage" to the wafer, thus the metrology must be considered as
an integral part of the process, with a value of some x% of the
value of the entire litho step. This is especially true as APC
enters into processing.
|
|
|
|
|
|
Report Abuse
|
|
|
Benjamin Bunday
Fellow
|
Comment : Per your yield excursion example, true, you are saving loss up
front (analogous to "playing defense"), but the increased
learning rate, if during product ramp up, can lead to better
yields in the better part of the price erosion curve, thus
creating value.
|
|
|
|
|
|
Report Abuse
|
|
|
Benjamin Bunday
Fellow
|
Comment : Respectful response to comments from Krystyna Collins, above:
Thanks for your comment; the purpose of this paper was to make
people think about changing their views on this topic, and open
discussions to help the community evolve their perceptions, and
thus your comments are very welcome as they spark this very
useful discussion. I respectfully disagree with a couple of your
points; please feel free to comment back if you wish.
|
|
|
|
|
|
Report Abuse
|
|
|
Krystyna Hutcheson
Member
|
Comment : Some of your adds are really loss prevention. For expl, when a
yld excur is caught by insp, it has not added any valu to the
wafer. The value is either lost to the excur or future loss is
being prevented by the que as the source is fixed. You make
money, but you don't add any value, like a litho tool does when
it adds a pattern. It is the additive idea that sets so many off
in the wrong direction. Just because it's not adding value
doesn't mean insp is not a valuable add to the process flow.
|
|
|
|
|
|
Report Abuse
|
|
|
Dan Hutcheson
Member
|
Comment : I've alway's answer the question as Value-add not the only
thing. Metrology & test are inherently value-assuring, as you can
burn a lot of money and brand value on bad product. But this
paper takes Metrology to the next level, where they truely do
have a value add. Very impressive work.
|
|
|
|
|
|
Report Abuse
|
|
|
|
|
|
|
|
|
|
|
|
Copyright © 2012 weSRCH. All rights reserved 
