IEM Seminar Series: On Low-Frequency Noise in Nanoscale Devices and CMOS Logic Gates
Event details
| Date | 02.02.2026 |
| Hour | 15:15 › 16:00 |
| Speaker | Léopold Van Brandt, UCLouvain, Belgium |
| Location | |
| Category | Conferences - Seminars |
| Event Language | English |
Abstract
The MOS transistors of smallest dimensions in advanced technology nodes are universally favoured for the design of digital integrated circuits but are very sensitive to all types of uncertainties. Among them, the intrinsic noise, a time-dependent variability, becomes an additional concern for the design of ultra-low power circuits and systems, as those operate at downscaled supply voltage. The seminar will begin with a brief didactic review of the mechanisms of charge carrier number and mobility fluctuations due to traps in semiconductors. The resulting current fluctuations over time are referred to as low-frequency noise. At UCLouvain, we measured random telegraph (or burst or popcorn) noise (RTN) in 28 nm FD-SOI transistors whose relative current variation can reach at least 30% in weak inversion. This experimentally confirmed the need to worry and to care about the possibly rare but proven worst cases in circuit simulation. Moreover, due the inherently stochastic nature of the defects, each device has a different variance for the observed random fluctuations of the threshold voltage over time. A collaboration with a Brazilian group (Universidade Federal do Rio Grande do Sul, Porto Alegre) of experts specialised in low-frequency noise modelling supported by NXP Semiconductors focused on RTN statistical modelling. The area scaling of this performance variability will be detailed, intuitively explained and discussed as part of the seminar. The published statistical model can furthermore be combined with the stochastic propagation principle to predict the timing variability in CMOS logic gates and ring oscillators. The accuracy of the predictions is assessed by comparison with transient noise simulations performed in a Monte-Carlo fashion with the implemented physics-based statistical RTN model.
Bio
Léopold Van Brandt was born in Belgium, in 1995. He received the B.S. and the M.S. degrees in Electrical Engineering from the Université catholique de Louvain, Louvain-la-Neuve, in 2015 and 2017, respectively, and the PhD degree in Engineering Science for his dissertation entitled “Statistical Analyses of Intrinsic Noise and Variability Effects in CMOS Digital Latches” in 2022. He then worked for two years as a postdoctoral research fellow in the Mathematical Engineering department of the UCLouvain on the project “Thermodynamics of Circuits for Computation”. Since October 2024, he has been leading his own postdoctoral research project entitled “Stochastic Modelling of Present and Future Nonlinear Dynamical Electronic Devices and Circuits”.
His research interests include but are not limited to: nanoelectronics; characterisation and modelling of the noise in nonlinear devices and circuits; circuit simulation theory; circuit reliability assessment (notably CMOS SRAM bitcells); stochastic thermodynamics; analog neuromorphic circuits and sensors, memristive and ferroelectric devices.
The interplay between research and education is one of his major concerns.
The MOS transistors of smallest dimensions in advanced technology nodes are universally favoured for the design of digital integrated circuits but are very sensitive to all types of uncertainties. Among them, the intrinsic noise, a time-dependent variability, becomes an additional concern for the design of ultra-low power circuits and systems, as those operate at downscaled supply voltage. The seminar will begin with a brief didactic review of the mechanisms of charge carrier number and mobility fluctuations due to traps in semiconductors. The resulting current fluctuations over time are referred to as low-frequency noise. At UCLouvain, we measured random telegraph (or burst or popcorn) noise (RTN) in 28 nm FD-SOI transistors whose relative current variation can reach at least 30% in weak inversion. This experimentally confirmed the need to worry and to care about the possibly rare but proven worst cases in circuit simulation. Moreover, due the inherently stochastic nature of the defects, each device has a different variance for the observed random fluctuations of the threshold voltage over time. A collaboration with a Brazilian group (Universidade Federal do Rio Grande do Sul, Porto Alegre) of experts specialised in low-frequency noise modelling supported by NXP Semiconductors focused on RTN statistical modelling. The area scaling of this performance variability will be detailed, intuitively explained and discussed as part of the seminar. The published statistical model can furthermore be combined with the stochastic propagation principle to predict the timing variability in CMOS logic gates and ring oscillators. The accuracy of the predictions is assessed by comparison with transient noise simulations performed in a Monte-Carlo fashion with the implemented physics-based statistical RTN model.
Bio
Léopold Van Brandt was born in Belgium, in 1995. He received the B.S. and the M.S. degrees in Electrical Engineering from the Université catholique de Louvain, Louvain-la-Neuve, in 2015 and 2017, respectively, and the PhD degree in Engineering Science for his dissertation entitled “Statistical Analyses of Intrinsic Noise and Variability Effects in CMOS Digital Latches” in 2022. He then worked for two years as a postdoctoral research fellow in the Mathematical Engineering department of the UCLouvain on the project “Thermodynamics of Circuits for Computation”. Since October 2024, he has been leading his own postdoctoral research project entitled “Stochastic Modelling of Present and Future Nonlinear Dynamical Electronic Devices and Circuits”.
His research interests include but are not limited to: nanoelectronics; characterisation and modelling of the noise in nonlinear devices and circuits; circuit simulation theory; circuit reliability assessment (notably CMOS SRAM bitcells); stochastic thermodynamics; analog neuromorphic circuits and sensors, memristive and ferroelectric devices.
The interplay between research and education is one of his major concerns.
Practical information
- General public
- Free
Contact
- Karin Jaymes