Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 20 of 20 for “"computing-in-memory"”.
-
Spiking Neural Network with Memristive Based Computing-In-Memory Circuits and Architecture
In recent years neuromorphic computing systems have achieved a lot of success due to its ability to process data much faster and using much less power compared to traditional Von Neumann computing architectures. There are two main types of Artificial Neural Networks (ANNs), Feedforward Neural …
-
Reliability Characterization of Mixed-Signal Circuits: Integrated Voltage Regulators and Computing-in-Memory Systems
As modern computing systems demand higher performance and energy efficiency, two emerging mixed-signal systems—Integrated Voltage Regulators (VRs) and Computing-in-Memory (CIM) systems—are increasingly vital in addressing power and data movement challenges. Integrated VRs provide fine-grained, …
-
Cryogenic Compute-In-Memory Circuits and Architectures for Spiking Neural Networks
This work presents a Computing-In-Memory (CIM)-based Spiking Neural Network (SNN) architecture for energy-efficient, low-latency inference in cryogenic conditions. The proposed CIM-SNN integrates current-domain 8T SRAM vector-matrix multiplication, cryogenic-aware temporal encoding, and …
-
Architecture and Circuit Design Optimization for Compute-In-Memory
… of the proposed research is to optimize computing-in-memory (CIM) design for accelerating Deep Neural Network (DNN) algorithms. As compute peripheries such as analog-to-digital converter (ADC) introduce significant overhead in CIM inference design, the research first focuses on the …
-
Enabling Energy-Efficient Hybrid CMOS and Embedded Memory Accelerators for Neuromorphic Computing at the Edge
The deployment of deep learning at the edge promises advances in autonomous driving, computer vision, and IoT, but is limited by the inefficiencies of conventional von Neumann architectures. The physical separation of memory and processing creates a performance bottleneck, with high energy and …
-
Automated Synthesis of Unconventional Computing Systems
Despite decades of advancements, modern computing systems which are based on the von Neumann architecture still carry its shortcomings. Moore's law, which had substantially masked the effects of the inherent memory-processor bottleneck of the von Neumann architecture, has slowed down due to …
-
Architecture optimization and design tools for CAM-based accelerators : from general-purpose to machine learning
The integration of AI/ML methods across various domains has been fueled by the vast amount of accumulated data over the past years. However, the energy-intensive nature of data-centric applications, compounded by the limitations of current computing models rooted in von Neumann’s design principle, …
-
Ab Initio Study of Interfaces in Halide Perovskites and Prediction of Two-Dimensional Ferroelectric Materials
Ab initio method is widely used as an approach for atomic-level understanding of mechanisms and a cost-effective tool for exploring target materials with advanced properties. This dissertation harnesses the strengths of ab initio method to investigate Metal Halide Perovskites (MHPs) and predict new …
-
Ultra-low Power Circuits and Architectures for Neuromorphic Computing Accelerators with Emerging TFETs and ReRAMs
Neuromorphic computing using post-CMOS technologies is gaining increasing popularity due to its promising potential to resolve the power constraints in Von-Neumann machine and its similarity to the operation of the real human brain. To design the ultra-low voltage and ultra-low power …
-
Algorithm-hardware co-optimization for neural network efficiency improvement
… neural networks (DNNs) are tremendously applied in the artificial intelligence field. While the performance of DNNs is continuously improved by more complicated and deeper structures, the feasibility of deployment on edge devices remains a critical problem. In this thesis, we present …
-
Otimização espaço temporal da lógica stateful memresistiva explorando as formas padrão de representação de expressões booleanas
A indústria de semicondutores vem enfrentando dificuldades crescentes para lidar com desafios tecnológicos e arquiteturais, como o esgotamento da miniaturização CMOS e as limitações de desempenho impostas pelo gargalo de von Neumann. Ao mesmo tempo, esse mercado precisa atender a demandas …
-
TWO-DIMENSIONAL MATERIALS-BASED SYNAPIC DEVICES FOR NEUROMORPHIC/IN-MEMORY COMPUTING APPLICATIONS
… (2D) materials-based synaptic devices are showing great potential to build neuromorphic/in-memory computing hardware. Their unique switching mechanisms and novel heterostructures provide new perspectives for synaptic device applications. This thesis aims to realize the 2D materials-based …
-
Enabling Full-Fledged Parallelism on Intermittently Powered Computing
Energy-harvesting batteryless devices exploit power from various sources, such as radio waves, sunlight, and vibration. However, the sporadic availability of ambient energy causes frequent power failures, forcing the systems to operate intermittently. The computation interruptions violate forward …
-
Non-volatile In-memory Computing for Large Scale Data-Intensive Workloads: Challenges and Opportunities
The application(domain)s that depend on the large amount of data for solving problems, e.g., genome sequence analysis, graph analytics, machine learning etc., suffer from growing overhead of data communication between physically separate logic (i.e., compute) and memory elements in conventional von …
-
Overcoming technological barriers in ReRAM based neural networks by addressing scalability and complexity challenges
As Artificial Intelligence techniques gain popularity, researchers are working to make Neural Networks (NNs) more energy-efficient, addressing challenges posed by limited device resources. One promising solution is using analog Resistive Random Access Memory (ReRAM) for Matrix-Vector Multiplication …
-
Robust Acceleration of Data-Centric Applications using Resistive Computing Systems
With the accessible data reaching zettabyte level, CMOS technology is reaching its limit for the data hungry applications. Moore's law has been reaching its depletion in recent studies. On the other hand, von Neumann architecture is approaching the bottleneck due to the data movement between the …
-
Approximate Computing and In-Memory Computing: the Best of the Two Worlds!
Machine learning (ML) has become ubiquitous, integrating into numerous real-life applications. However, meeting the computational demands of ML systems is challenging, as existing computing platforms are constrained by memory bandwidth, and technology scaling no longer yields substantial …
-
High Performance Compute Accelerators for Radio Frequency Signal Processing Applications
Signal processing algorithms for Radio Frequency (RF) sensors, RF communication and RF interaction have seen remarkable progress, aligning with the rapid advancements in wireless communication standards. The resulting surge in data volume presents a challenge: processing this data at high …
-
Efficient and Reliable In-Memory Computing Systems Using Emerging Technologies: From Device to Algorithm
Big data applications are memory-intensive, and the cost of bringing data from the memory to the processor involves large overheads in energy and processing time. This has driven the push towards specialized accelerator units that can perform computations close to where the data is stored. Two …