Mmastodon TechnologySemiconductors first seen 24 d ago, last 24 d ago, peak #6
TUM researchers use neural networks to speed 3nm transistor simulation
Original: Researchers at Technical University of Munich developed neural-network transistor models that predict drain current and
Researchers at the Technical University of Munich have developed neural-network models of transistors that predict drain current and charges for a 3 nm nanosheet device, reported by Semiconductor Engineering. The approach aims to speed up SPICE circuit simulation, which becomes slow and computationally heavy at advanced process nodes, offering chip designers faster modelling of next-generation semiconductor behaviour.
Why now: Faster and more accurate simulation of 3 nm transistors matters to chip designers as process nodes shrink and conventional SPICE modelling struggles.
Technical University of MunichSemiconductor Engineeringnanosheet transistorSPICE simulation
Evidence
API: https://socialmediatrends-api.osmike.com/v1/trends/3663384