Jaro De Roose

Jaro De Roose

Auto-generated profile, based on publicly available information

Jaro De Roose

Department of Electrical Engineering, MICAS, KU Leuven, Leuven, Belgium

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
20 January 2023 in IEEE Journal of Solid-State Circuits
Verified
Journal Article
Closed access
VJVikram Jain
Vikram Jain
SGSebastian Giraldo
Sebastian Giraldo
JRJaro De Roose
Jaro De Roose
LMLinyan Mei
Linyan Mei
BBBert Boons
Bert Boons
MVMarian Verhelst
Marian Verhelst
Extreme edge devices or Internet-of-Things (IoT) nodes require both ultra-low power (ULP) always-on (AON) processing as well as the ability to do on-demand sampling and processing. Moreover, support for IoT applications, such as voice recognition, machine monitoring, and so on, requires the ability to execute a wide range of machine learning (ML) workloads. This brings challenges in hardware (HW) design to build flexible processors operating in ULP regime. This article presents TinyVers, a tiny versatile ULP ML system-on-chip (SoC) to enable enhanced intelligence at the extreme edge. TinyVers exploits dataflow reconfiguration to enable multi-modal support and aggressive on-chip power management for duty cycling to enable smart sensing applications. The SoC combines an reduced instruction set computer-V (RISC-V) host processor, a 17-tera operations per second per watt (TOPS/W) dataflow reconfigurable ML accelerator, a 1.7- $\mu$ W deep sleep wake-up controller (WuC), and an embedded magnetoresistive random access memory (eMRAM) for boot code and ML parameter retention. The SoC can perform up to 17.6 giga operations per second (GOPS) while achieving a power consumption range from 1.7 $\mu$ W to 20 mW. Multiple ML workloads aimed for diverse applications are mapped on the SoC to showcase its flexibility and efficiency. All the models achieve 1–2 TOPS/W of energy efficiency with a power consumption below 230 $\mu$ W in continuous operation. In a duty-cycling use case for machine monitoring, this power is reduced to below 10 $\mu$ W.

20 August 2020 in IEEE Solid-State Circuits Letters
Verified
Journal Article
Closed access
JRJaro De Roose
Jaro De Roose
HXHaoming Xin
Haoming Xin
AHAhmed Hallawa
Ahmed Hallawa
GAGerd Ascheid
Gerd Ascheid
PHPieter J. A. Harpe
Pieter J. A. Harpe
MVMarian Verhelst
Marian Verhelst
We present a 5-sensor, fully-integrated sensing system with inter-changeable sensors and programmable configuration to create a sub-microWatt multi-sensor node that can tackle a wide range of sensing applications. Furthermore, the sensor node is capable of autonomously adapting its configuration to the application requirements hence minimizing system power. Such self-reconfiguration is enabled at low overhead by developing an automated offline optimization strategy, in combination with an autonomous embedded configuration controller, using the concept of Behavioral Trees (BT). The resulting fully-integrated platform consumes a maximum of 321nW when sampling at 500Hz and 3025nW at 8kHz. Furthermore we demonstrate the end-to-end autonomous optimization flow for two different applications exploiting different sensors: Human Activity Recognition using accelerometers and Machine Listening using a microphone. Both use cases demonstrate that the introduced system and methodology reduces the power by more than a factor 2 without losing significant application detection accuracy.