The 17th International Conference on PhD Research in Microelectronics and Electronics (PRIME) will take place from 12 to 15 June 2022 in Sardinia (Italy).
Two papers from our StorAIge partners will be presented.
- One paper from from University of Bologna and ST Microlectonics Italy:
Phase-change memory cells characterization in an analog in-memory computing perspective
Alessio Antolini, Andrea Lico, Eleonora Franchi Scarselli, Marcella Carissimi, Marco Pasotti
Abstract: Power consumption related to data transfers between processing and memory units has become a critical issue in the recent data-centric outlook of integrated circuits. In the context of In-memory Computing (IMC), where data conveyance is narrowed performing computations directly within the memory unit, Phase-change Memory (PCM) technology has become an attractive candidate due to its intrinsic multilevel storage capability. The test vehicle of this work is an embedded PCM (ePCM) provided by STMicroelectronics and designed in 90-nm smart power BCD technology with a Ge-rich Ge-Sb-Te (GST) alloy for automotive applications. In this framework, a preliminary characterization of PCM cells has been carried out, aimed at evaluating their performance as enabling devices for analog in-memory computing (AIMC) applications.
- One paper from Politecnico di Torino:
Energy-efficient and Privacy-aware Social Distance Monitoring with Low-resolution Infrared Sensors and Adaptive Inference
Chen Xie, Daniele Jahier Pagliari, Andrea Calimera
Abstract: Low-resolution infrared (IR) Sensors combined with machine learning (ML) can be leveraged to implement privacypreserving social distance monitoring solutions in indoor spaces. However, the need of executing these applications on Internet of Things (IoT) edge nodes makes energy consumption critical. In this work, we propose an energy-efficient adaptive inference solution consisting of the cascade of a simple wake-up trigger and a 8-bit quantized Convolutional Neural Network (CNN), which is only invoked for difficult-to-classify frames. Deploying such adaptive system on a IoT Microcontroller, we show that, when processing the output of a 8×8 low-resolution IR sensor, we are able to reduce the energy consumption by 37-57% with respect to a static CNN-based approach, with an accuracy drop of less than 2% (83% balanced accuracy).
More information : http://prime-conference.org/
