2005
DOI: 10.1109/tvlsi.2005.850106
|View full text |Cite
|
Sign up to set email alerts
|

Accurate prediction of substrate parasitics in heavily doped CMOS processes using a calibrated boundary element solver

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

0
2
0
2

Year Published

2006
2006
2015
2015

Publication Types

Select...
4
2

Relationship

0
6

Authors

Journals

citations
Cited by 6 publications
(4 citation statements)
references
References 16 publications
0
2
0
2
Order By: Relevance
“…2(b). The substrate can be modeled by a two port ʌ-resistance network [7]. Furthermore, other models exist, due to exact occasion of the modeled structure, such as an RLCG two parallel transmission interconnect line model or analytical models such as quasi-static (QS) and magnetic potential (MP) models [8][9].…”
Section: Substrate Structuresmentioning
confidence: 99%
“…2(b). The substrate can be modeled by a two port ʌ-resistance network [7]. Furthermore, other models exist, due to exact occasion of the modeled structure, such as an RLCG two parallel transmission interconnect line model or analytical models such as quasi-static (QS) and magnetic potential (MP) models [8][9].…”
Section: Substrate Structuresmentioning
confidence: 99%
“…A three layer approximation is used for heavily doped processes and a two layer approximation is used for lightly doped processes. In the calibration routine, Z 11 measurements for isolated contacts are compared with the results from the parasitic extractor and the layered doping profile is adjusted by an optimization algorithm until EPIC simulations are in good agreement with measurements [8]. The results of the calibration are sensitive to the initial state, so for the calibration to work successfully, a good initial estimate of the layered doping profile is required.…”
Section: Substrate Model Parameter Extractionmentioning
confidence: 99%
“…Συγκεκριμένα, το σφάλμα του πυρήνα ΓΑΣ κυμαίνεται κατά μέσο όρο στο 1,5%, ενώ για τον πυρήνα ΗΑΔ το μέσο σφάλμα είναι μικρότερο του 1%, και σε ορισμένες περιπτώσεις μάλιστα μειώνεται πρακτικά στο μηδέν. ΠΙΝΑΚΑΣ 5-6 Συγκριτικά αποτελέσματα για το χρόνο εκτέλεσης μεταξύ της προτεινόμενης μεθόδου και της [81]…”
Section: αποτελέσματα μετρήσεωνunclassified
“…από την[81] και για τους δυο πυρήνες. Επιπρόσθετα, ο πυρήνας ΓΑΣ είναι κατά ένα μικρό συντελεστή ταχύτερος από τον ΗΑΔ, όπως είναι αναμενόμενο, λόγω της απλούστερης μορφής των εξισώσεών του (αφού ο ΗΑΔ εκφράζεται ως ένα αρκετά μεγάλο άθροισμαάπειρο-όρων).…”
unclassified