High Predictability Global Routing during Floorplanning of Complex Chips

High Predictability Global Routing during Floorplanning of Complex Chips

Description

This project addresses the challenge of premature global routing in very large scale integrated circuit (VLSI) design. The goal is to improve the predictability and performance of routing immediately after the floorplanning stage and before final placement, using a mathematical optimization model based on multi-commodity flow and Linear Programming (LP). This methodology allows to minimize the total wiring length, the use of vias and the module crossings, while respecting the physical constraints of complex designs. The research integrates the theoretical rigor of Eindhoven University of Technology (TU/e) and the UPC with the practical validation of Qualcomm engineers, directly contributing to the development of state-of-the-art Electronic Design Automation (EDA) tools.

Background

Antoni graduated in Mathematics with a minor in Computer Science from the University of Barcelona (UB). He subsequently completed a Master's degree in Data Science and Artificial Intelligence through the EIT Digital program, with stays at the Eindhoven University of Technology (TU/e) and the University of Turku (UTU). His academic background stands out for its solid experience in optimization, graph modeling and algorithm design. Professionally, he has consolidated this knowledge by collaborating with Qualcomm and the UPC Chip Chair, where he has acquired practical experience in VLSI design flows and EDA tools.

Motivation

His research is driven by the exponential growth in chip complexity, which requires algorithms capable of anticipating routing problems early in the design flow. Antoni seeks to transform mathematical theory into practical engineering methods that improve design quality and reduce computational cost, generating real industrial impact. He is particularly motivated by the development of more reliable and sustainable EDA tools that accelerate innovation in the semiconductor industry.

Investigador/a de Suport a la Recerca

Antoni Pech Alberich

Antoni Pech Alberich

Degree in Mathematics and Master's in Data Science and Artificial Intelligence

Host Organization

Supervisors

Jordi Cortadella

Jordi Cortadella

UPC Supervisor

The content of this website reflects only the views of the Catedra Chip Chair UPC project.

Optimizing distributed AI workload training on multinode computing systems

Sergi Tomàs Martínez

Sergi Tomàs Martínez

Investigador/a de Suport a la Recerca

Simulation-Based Recommendation Framework for Scalable Training of Distributed AI

Tomàs Gadea Alcaide

Tomàs Gadea Alcaide

Investigador/a de Suport a la Recerca

Hierarchical floorplanning optimization algorithms for System-on-Chip architectures

Bernat Ibañez

Bernat Ibañez

Investigador/a de Suport a la Recerca

Mathematical Optimization Techniques for Hierarchical Floorplanning of Complex Chips

Yilihamujiang Yimamu

Yilihamujiang Yimamu

Investigador/a Predoctoral

Optimization of Training Distributed AI Workloads on Multi-node Computing Systems.

Xavier Querol Bassols

Xavier Querol Bassols

Investigador/a de Suport a la Recerca

Development of an Interactive Graphical Tool for Optimization and Editing of Floorplanning in Chip Design

Nuria Elizondo Cereza

Nuria Elizondo Cereza

Investigador/a de Suport a la Recerca

Optimization of Training Distributed AI Workloads on Multi-node Computing Systems

Mohammad Nasser

Mohammad Nasser

Investigador/a Predoctoral

Development of Mathematical and Heuristic Optimization Tools for Chip Floorplanning

Guillem Pastor Rué

Guillem Pastor Rué

Investigador/a de Suport a la Recerca

This site is registered on wpml.org as a development site. Switch to a production site key to remove this banner.