Algorithms and Techniques for Electronic Design Automation

For VLSI (Very Large Scale Integration) projects, computational tools are required to automate the design flow of integrated circuits. At ECL, we work with algorithms and techniques for chip design automation (EDA – Electronic Design Automation), investigating the integration of tools for approximate synthesis with traditional flows and enabling the reduction of energy consumption in various applications.

We explore the use of various machine learning techniques to improve the quality of solutions and reduce design time. Algorithms with a mathematical optimization method are also developed to position and route cells in the same step. These algorithms seek to reduce interconnections and thus increase the energy efficiency of devices.


Error-Tolerant Circuits and Systems

With the miniaturization of transistors, more and more particles with low electrical charge, coming from the sun or outside the galaxy, are capable of interacting and interfering with the operation of electronic devices. At ECL, we develop techniques to study and mitigate the effects of the interaction of electrical particles in different types of circuits and in different technological nodes.

pasted image 0

Work from ECL accepted at the International Symposium on Physical Design conference

The work titled “How Deep Learning Can Drive Physical Synthesis Towards More Predictable Legalization,” developed by Renan Netto, Sheiny Fabre, Tiago Augusto Fontana, Vinicius Livramento, Laércio Pilla, and José Luís Güntzel, was accepted for publication at the International Symposium on Physical Design event in 2019, being one of the candidates for the best paper award at the same event.

This work deals with the use of neural networks to identify regions with high cell movement during the legalization of integrated circuits.

The International Symposium on Physical Design is one of the main conferences in the field of Electronic Design Automation (EDA) and physical design.


Work from ECL accepted on Symposium on Circuits and Systems international conference

The work entitled “Design of Energy-efficient Gaussian Filters by Combining Refactoring and Approximate Adders,” developed by Marcio Monteiro, Pedro Aquino, Ismael Seidel, Mateus Grellert, Leonardo Soares, José Luis Güntzel, and Cristina Meinhardt, has been accepted for publication at the International Symposium on Circuits and Systems.

This work deals with the combination of logical refactoring and the use of approximate computing in an architecture for Gaussian filter to optimize area and energy.

The International Symposium on Circuits and Systems is one of the leading conferences in the field of circuits and systems.



Keynote Talk:

Speaker: Prof. David Patterson, University of California at Berkeley and Google, USA

August 25, 2021, 13:00

Title: Ten Lessons From Three Generations Shaped Google’s TPUv4i