Abstract:With shrinking interconnect spacing in advanced technology nodes, existing timing predictions become less precise due to the challenging quantification of crosstalk-induced delay. During the routing, the crosstalk effect is typically modeled by predicting coupling capacitance with congestion information. However, the timing estimation tends to be overly pessimistic, as the crosstalk-induced delay depends not only on the coupling capacitance but also on the signal arrival time. This work presents a crosstalk-aware timing estimation method using a two-step machine learning approach. Interconnects that are physically adjacent and overlap in signal timing windows are filtered first. Crosstalk delay is predicted by integrating physical topology and timing features without relying on post-routing results and the parasitic extraction. Experimental results show a match rate of over 99% for identifying crosstalk nets compared to the commercial tool on the OpenCores benchmarks, with prediction results being more accurate than those of other state-of-the-art methods.

What problem does this paper attempt to address?

This paper attempts to solve the problem of the decline in the accuracy of existing timing predictions caused by the reduction in the spacing between interconnect lines at advanced process nodes, especially due to the difficulty in quantifying the delay caused by coupling (i.e., crosstalk delay). During the routing process, crosstalk effects are usually modeled by predicting the coupling capacitance, but such predictions are often too pessimistic because the crosstalk - induced delay depends not only on the coupling capacitance but also on the signal arrival time. This paper proposes a crosstalk - aware timing estimation technique based on a two - step machine - learning method, aiming to improve the accuracy of timing predictions, reduce the number of design iterations, and thus accelerate the timing convergence process. Specifically, the main contributions of the paper include: 1. **Introducing a new crosstalk - aware delay prediction method**: The relative arrival time between the aggressor and the victim is extracted and its impact is quantified to achieve accurate crosstalk - induced delay estimation. Physically adjacent interconnect lines that do not overlap in the timing window are filtered out, avoiding prediction complexity and pessimism. 2. **Proposing a two - step machine - learning model**: This model integrates physically - related features and timing - related features and can capture crosstalk - induced delays without relying on post - routing netlists and parasitic parameter information. The model is flexible and can be integrated into existing routing optimization tools, showing broad application potential. Through these methods, the paper aims to provide a more accurate and efficient timing prediction solution, especially in high - density design scenarios, which can significantly improve prediction accuracy, reduce the number of design iterations, and accelerate the chip design process.

A Crosstalk-Aware Timing Prediction Method in Routing

Crosstalk Prediction of Single- and Double-Walled Carbon-Nanotube (SWCNT/DWCNT) Bundle Interconnects

A Coupling and Crosstalk Considered Timing-Driven Global Routing Algorithm for High Performance Circuit Design

Priority-Based Routing Resource Assignment Considering Crosstalk.

A Cross-Layer Framework for Temporal Power and Supply Noise Prediction

Post Global Routing Crosstalk Risk Estimation and Reduction

Crosstalk-Aware Routing Resource Assignment

Machine Learning Based Routing Congestion Prediction in FPGA High-Level Synthesis

A novel method for worst-case interconnect delay estimation

Crosstalk Driven Routing Resource Assignment

Estimation Methods of Interconnect Delay and Crosstalk

ETEM: an efficient gate and interconnect timing estimator considering cross-coupling for high performance layout

Crosstalk and Congestion Driven Layer Assignment Algorithm

Power/Ground Network Aware and Row-Based Solutions to the Crosstalk Driven Routing Problem

New Crosstalk Avoidance Codes Based on a Novel Pattern Classification

Prediction-based routing for real time communications in wireless multi-hop networks.

A Novel Fine-Grain Track Routing Approach For Routability And Crosstalk Optimization

MSS: Exploiting Mapping Score for CQF Start Time Planning in Time-Sensitive Networking

Accurate Prediction of Detailed Routing Congestion Using Supervised Data Learning.

Layer Assignment Algorithm For Rlc Crosstalk Minimization

Dynamic Routing, Wavelength and Core Allocation in Multi-Core Fiber Based Optical Networks Considering Inter-Core Crosstalk