A5054580844- Energy Harvesting in Wireless Networks
- Innovative Energy Harvesting Technologies
- Parallel Computing and Optimization Techniques
- Advanced Memory and Neural Computing
- Low-power high-performance VLSI design
- Semiconductor materials and devices
- Wireless Power Transfer Systems
- Advancements in Semiconductor Devices and Circuit Design
- Interconnection Networks and Systems
- Energy Efficient Wireless Sensor Networks
- Ferroelectric and Negative Capacitance Devices
- Microgrid Control and Optimization
- Embedded Systems Design Techniques
- Advanced DC-DC Converters
- Physical Unclonable Functions (PUFs) and Hardware Security
- IoT and Edge/Fog Computing
- Advanced Data Processing Techniques
- VLSI and FPGA Design Techniques
- Neuroscience and Neural Engineering
- Smart Grid Energy Management
- Advanced Sensor and Energy Harvesting Materials
- Analog and Mixed-Signal Circuit Design
- Stochastic Gradient Optimization Techniques
- Security and Verification in Computing
- Advanced Battery Technologies Research
CEA Grenoble
2011-2024
Commissariat à l'Énergie Atomique et aux Énergies Alternatives
2012-2024
Université Grenoble Alpes
2010-2024
CEA LIST
2021
Integra (United States)
2021
CEA LETI
2010-2020
Institut polytechnique de Grenoble
2010-2019
Increased capabilities such as recognition and self-adaptability are now required from IoT applications. While node power consumption is a major concern for these applications, cloud-based processing becoming unsustainable due to continuous sensor or image data transmission over the wireless network. Thus optimized ML transfers should be integrated in node. Moreover, applications torn between sporadic data-logging energy-hungry (e.g. classification). Thus, versatility of key addressing this...
In this paper we present a methodology allowing an emulated-3D two tiers physical implementation of any design using 2D commercial tools. Place and Route is achieved through similar steps as required by designs: pre clock tree synthesis (including placement), routing; to which added folding step in order emulate the 3D placement. Routing both parallel inter-tier metal layers made possible modifying input files Our study covers power supply network on tiers, forbidden via active placement...
IoT node application requirements are torn between sporadic data-logging and energy-hungry data processing (e.g. image classification). This paper presents a versatile covering this gap in energy by leveraging two on-chip sub-systems: low power, clock-less, event-driven Always-Responsive (AR) part an energy-efficient On-Demand (OD) part. The AR contains 1.7MOPS event-driven, asynchronous Wake-up Controller (WuC) with 207ns wake-up time optimized for short computing. OD combines deep-sleep...
Asynchronous circuits have characteristics that differ significantly from those of synchronous in terms their power and robustness to variations. In this article, we show how it is possible exploit these design robust ultralow-power within the scope Internet Everything (IoE) with globally asynchronous locally (GALS) architectures. More specifically, our aim describe fundamentals circuit design; detail specific methodologies practical examples low-power, circuits; offer clear guidelines...
Autonomous devices that are self-powered by extracting their energy from environment a new opportunity for monitoring purposes. A multi-energy sources and multi-sensors microsystem targeting autonomous wireless sensor node applications is presented. Since the available not constant over time due to very low harvested power levels, an efficient management strategy mandatory. In this paper, we propose use benefits of event-based asynchronous controller reduce overall consumption bring voltage...
In the context of highly data-centric applications, close reconciliation computation and storage should significantly reduce energy-consuming process data movement. This letter proposes a computational SRAM (C-SRAM) combining in- near-memory computing (IMC/NMC) approaches to be used by scalar processor as an energy-efficient vector processing unit. Parallel is thus performed on vectorized integer large words using usual logic arithmetic operators. Furthermore, multiple rows can...
Editors' note:Asynchronous circuits are well-suited to ultra-low-power design. This article presents a microsystem that is powered only by energy extracted from the environment implement an autonomous sensing application. Key this application use of asynchronous logic, which not provides greater efficiency due its event-driven nature but, more importantly, allows graceful adaptation highly variable power availability.—Montek Singh (UNC Chapel Hill) and Luciano Lavagno (Politecnico di Torino)
Wireless sensor nodes that are self-powered by extracting their energy from environment a new opportunity for monitoring purpose. Since the available is not constant over time and due to very low harvested power levels, efficient management strategies mandatory improving autonomy. At system level, scheduling algorithms proposed efficiently use multi path architectures avoid as much possible of batteries. A data- energy-driven architecture its associated algorithm presented achieving high...
This paper provides a comprehensive review of Three-Independent-Gate Field-Effect Transistors (TIGFETs). In parallel to the focus on transistor scaling, an alternative approach push further performance computing systems consists in increasing functionalities basic transistors by means additional gate controls. TIGFETs belong this category devices and can achieve different modes operation according bias terminals. particular, these are capable (i) device-level polarity control, (ii) dynamic...
Extracting energy from the environment (e.g. light, heat, vibrations or radiations) can be useful to enhance microsystems lifespan and even reach autonomy. In this paper, we propose an innovative Energy Harvesting Platform (EHP) architecture which is used in design of a multi-energy sources multi-sensors microsystem suitable for Wireless Sensor Nodes (WSN) applications such as ambient intelligence monitoring purposes. Through use additional direct power path sensors dedicated low...
This letter presents a single-rail two-port static random-access memory (SRAM) designed in 28-nm FD-SOI technology specifically for synchronous/asynchronous Internet of Things node. SRAM supports an asynchronous interface communication and fast transition sleep/active mode. It enables simultaneous synchronous accesses on its two independent ports, as well selective virtual ground (SVGND) to support ultralow voltage read operations. Measurements 64-kb macro prove the functionality supply...
In this work, we explore the benefits of Fully Depleted Silicon-on-Insulator (FDSOI) devices to implement full-adders focused operate at sub-threshold level. Their design based on asynchronous proposal shows lower energy consumption than their synchronous counterparts. Afterwards, use FDSOI back-gate bias (V BB ) allows an enhancement circuit behavior, in contrast other technology proposals. proper management could provide a temperature insensitive behavior and larger variability robustness.
Achieving high energy efficiency harvesting platforms requires tracking variations of the levels. Leveraging storage components whose voltage level varies with state charge, it becomes efficient to perform monitoring. In this paper, we propose two types analog-to-digital monitoring interfaces. both cases, their outputs directly fit asynchronous 4-phases protocol and Quasi Delay Insensitive (QDI) logic. On one hand, in a passive scheme, platform waits for energy-events. Reacting threshold...
Due to low activity in Internet of Things (IoT) applications, systems tend leverage power modes order reduce their consumption. Normally-off computing thus arose, consisting having turned off most part a system’s supply, while dynamically turning on components as the application needs it. As wake up sources may be diverse, simple controllers are integrated handle smart schemes. Therefore, prevent overconsumption transitioning running mode, fast sequences required. An asynchronous 16-bit...
In energy-harvesting Internet of Things (EH-IoT) wireless networks, maintaining energy neutral operation (ENO) is crucial for their perpetual and maintenance-free property. Guaranteeing this ENO condition optimal power-performance trade-off under transient harvested channel quality particularly challenging. This paper proposes a multi-agent actor-critic reinforcement learning modulating both the transmitter duty-cycle output power based on state-of-buffer (SoB) state-of-charge (SoC)...
Ultra Thin Body and Box Fully Depleted Silicon-on-Insulator (UTBB FDSOI) devices are promising candidates to substitute planar CMOS technology for very large-scale integrated (VLSI) circuits, recently proved implement complex systems [1]. Their main characteristic is the use of Buried-Oxide (BOX) by providing high channel isolation excellent electrostatic control over channel. Due BOX feasible combine n-/p-type backplane, allowing different threshold voltages (V T ) same gate stack, e.g....
Reinforcement learning (RL) has received much attention in recent years due to its adaptability unpredictable events such as harvested energy and workload, especially the context of edge computing for Internet-of-Things (IoT) nodes. Due limited resources IoT nodes, it is difficult achieve self-adaptability. This paper studies online reactivity issues fixed rate linear actor-critic (LAC) algorithm transmission duty-cycle control. We propose LAC-AB that introduces into LAC an adaptive called...
On-chip memories, and in particular SRAMs, are among the most critical components terms of data security because they might contain sensitive such as secret keys. Whenever a tampering event is detected, one should be able to erase efficiently rapidly full content memory holding data, but current solutions based on simple power-off lead very long erasure times. In this paper, we present non-invasive design technique an innovative mechanism remove electric charges from SRAM bitcells still...