Estimation of interconnect parasitic capacitance in an open-source VLSI design flow

This work investigates the possibility of early estimation of the parasitic capacitance of interconnects in digital VLSI blocks using features available after clock tree synthesis but before detailed routing. A fully open-source design flow based on Yosys, OpenROAD, and OpenRCX and using the Sky130 standard-cell library was developed to generate the dataset. The dataset contains 40 RTL blocks, 638 physical implementations, and more than 8 million interconnects. Linear regression models and XGBoost gradient boosting models using geometrical, topological, load-related, and global implementation features were investigated. The models were evaluated using five-fold group cross-validation with splitting by RTL block. The best result was obtained by the XGBoost model: the mean absolute error was 0.58 fF, WAPE was 15.4 %, and the coefficient of determination R2 was 0.94. Compared with the linear HPWL-based model, the mean absolute error was reduced by 35.9 %. The obtained results confirm the possibility of accurate estimation of post-route parasitic capacitance within a fully open and reproducible VLSI design flow.

Authors

Institutions

Publication Details

Journal
INFORMACIONNYE TEHNOLOGII
Published
2026-11-20
DOI
https://doi.org/10.17587/it.32.582-589
Citations
1
Primary Topic
VLSI and FPGA Design Techniques
Type
article
Field-Weighted Citation Impact
1.03
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Estimation of interconnect parasitic capacitance in an open-source VLSI design flow

N. V. Zheludkov, S. I. Aryashev, Y. A. Kulemina
1 citations
INFORMACIONNYE TEHNOLOGII
VLSI and FPGA Design Techniques
1.03
article

Estimation of interconnect parasitic capacitance in an open-source VLSI design flow

N. V. Zheludkov, S. I. Aryashev, Y. A. Kulemina
article en
1 citations

Abstract

This work investigates the possibility of early estimation of the parasitic capacitance of interconnects in digital VLSI blocks using features available after clock tree synthesis but before detailed routing. A fully open-source design flow based on Yosys, OpenROAD, and OpenRCX and using the Sky130 standard-cell library was developed to generate the dataset. The dataset contains 40 RTL blocks, 638 physical implementations, and more than 8 million interconnects. Linear regression models and XGBoost gradient boosting models using geometrical, topological, load-related, and global implementation features were investigated. The models were evaluated using five-fold group cross-validation with splitting by RTL block. The best result was obtained by the XGBoost model: the mean absolute error was 0.58 fF, WAPE was 15.4 %, and the coefficient of determination R2 was 0.94. Compared with the linear HPWL-based model, the mean absolute error was reduced by 35.9 %. The obtained results confirm the possibility of accurate estimation of post-route parasitic capacitance within a fully open and reproducible VLSI design flow.

INFORMACIONNYE TEHNOLOGIIVol. 32(11)
Kurchatov Institute (RU)
Industry, innovation and infrastructure
Openalex Percentile: Top 30%
VLSI and FPGA Design Techniques
1.03
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon