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 60 for “"compute in-memory"”.

  1. Compute-In-Memory Deep Learning Using Memristor Crossbars

    Deep learning redefines computing and greatly impacts our daily lives, but it comes at the cost of massive energy consumption and intensive memory requirements from large-scale data centers driven by exploding data collection and artificial intelligence. This challenge demands an intense search for …

    tamu Repository record for Compute-In-Memory Deep Learning Using Memristor Crossbars (opens in a new tab) Embargoed until 2027-05-01

  2. 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 circuit …

    gatech Repository record for Architecture and Circuit Design Optimization for Compute-In-Memory (opens in a new tab)

  3. Efficient and Robust Compute-in-Memory for Edge Intelligence

    With deep neural networks (DNNs) fueling major advancements in artificial intelligence (AI), the demand for efficient computing solutions has never been higher. This dissertation delves into the challenges and innovations in the field of compute-in-memory (CIM) technologies, a key area for …

    gatech Repository record for Efficient and Robust Compute-in-Memory for Edge Intelligence (opens in a new tab)

  4. COMPUTE-IN-MEMORY ON EMERGING MEMORY TECHNOLOGY: FROM DEVICE TO ALGORITHM

    Present-day computing systems largely adhere to the von Neumann architecture, which necessitates data transfer from memory to a processing unit. This architecture presents a significant latency issue when accessing data from memory units, which is a primary performance hindrance for various …

    penn Repository record for COMPUTE-IN-MEMORY ON EMERGING MEMORY TECHNOLOGY: FROM DEVICE TO ALGORITHM (opens in a new tab)

  5. 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 …

    gatech Repository record for Cryogenic Compute-In-Memory Circuits and Architectures for Spiking Neural Networks (opens in a new tab)

  6. Benchmarking Test-Time DNN Adaptation at Edge with Compute-In-Memory

    … the edge can deteriorate over time due to shifts in the data distribution. For heightened robustness, it’s crucial for DNNs to continually refine and improve their predictive capabilities. However, adaptation in resource-limited edge environments is fraught with challenges: (i) new labeled data …

    gatech Repository record for Benchmarking Test-Time DNN Adaptation at Edge with Compute-In-Memory (opens in a new tab)

  7. DURACIM: Durable compute-in-memory (CiM) inference for resource-limited environments

    … encounter a performance bottleneck known as the memory wall, where the compute units are faster and consume less energy than data movement from memory. Compute-in-Memory (CIM) architectures can support MAC operations within memory arrays, helping to resolve the memory wall. However, they suffer …

    iastate Repository record for DURACIM: Durable compute-in-memory (CiM) inference for resource-limited environments (opens in a new tab)

  8. Reliability and Security of Compute-In-Memory Based Deep Neural Network Accelerators

    Compute-In-Memory (CIM) is a promising solution for accelerating DNNs at edge devices, utilizing mixed-signal computations. However, it requires more cross-layer designs from algorithm levels to hardware implementations as it behaves differently from the pure digital system. On one side, the …

    gatech Repository record for Reliability and Security of Compute-In-Memory Based Deep Neural Network Accelerators (opens in a new tab)

  9. Compute-in-memory designs for deep neural network and combinatorial optimization problems accelerators

    The unprecedented growth in Deep Neural Networks (DNN) model size has resulted into a massive amount of data movement from off-chip memory to on-chip processing cores in modern Machine Learning (ML) accelerators. Compute-In-Memory (CIM) designs performing analog DNN computations within a memory …

    texas Repository record for Compute-in-memory designs for deep neural network and combinatorial optimization problems accelerators (opens in a new tab)

  10. COMPUTE-IN-MEMORY WITH EMERGING NON-VOLATILE MEMORIES FOR ACCELERATING DEEP NEURAL NETWORKS

    … accelerate deep neural networks (DNNs) with emerging non-volatile memories (eNVMs) based compute-in-memory (CIM) architecture. The research first focuses on the inference acceleration and proposes a resistive random access memory (RRAM) based CIM architecture. Two generations of RRAM testchips …

    gatech Repository record for COMPUTE-IN-MEMORY WITH EMERGING NON-VOLATILE MEMORIES FOR ACCELERATING DEEP NEURAL NETWORKS (opens in a new tab)

  11. Design of a non-volatile capacitor-based compute-in memory neural network accelerator

    To avoid growing challenges with existing neural network accelerators which have sig- nificant energy requirements, the design for a compute-in-memory (CIM) chip is presented and demonstrated in simulation. The CIM aspect is based off a ferroelectric memory de- vice which allows for both storing …

    gatech Repository record for Design of a non-volatile capacitor-based compute-in memory neural network accelerator (opens in a new tab)

  12. Overcoming Device Level Challenges to Enable Compute In-Memory with Emerging Non-Volatile Memory

    … research is to overcome challenges faced by compute in-memory (CIM) and embedded non-volatile memory (eNVM). The combination of eNVM and CIM can greatly increase memory bandwidth and memory density, perform computation on the bitline of memory sub-arrays, and minimize data transport. Despite …

    gatech Repository record for Overcoming Device Level Challenges to Enable Compute In-Memory with Emerging Non-Volatile Memory (opens in a new tab)

  13. Compute-in-memory for Efficient AI: Cross-layer Modeling and Hardware-software Co-Design

    Modern deep learning is bottlenecked by the memory wall, where moving weights and activations consumes several times the energy of doing computations. Compute-in-memory (CIM) is one of the main ways around the wall; by performing multiply-accumulate operations inside the memory arrays that store …

    gatech Repository record for Compute-in-memory for Efficient AI: Cross-layer Modeling and Hardware-software Co-Design (opens in a new tab)

  14. Design and calibration of current-mode compute-in-memory integrated circuits for neural network processing

    The growing use of artificial intelligence (AI) in edge devices has increased the need for hardware that can run neural network inference with higher energy efficiency than conventional digital processors. Compute-in-memory (CIM) addresses this challenge by reducing data movement and performing the …

    aalto Repository record for Design and calibration of current-mode compute-in-memory integrated circuits for neural network processing (opens in a new tab)

  15. An energy efficient compute-in-memory SRAM for low power CNN based machine learning application

    With the increase in computational parallelism and low-power Integrated Circuits (ICs) design, neuromorphic technologies and machine learning algorithms have returned to spotlight as practical solutions for implementing complex classification problems. Although GPUs have significantly advanced the …

    texas Repository record for An energy efficient compute-in-memory SRAM for low power CNN based machine learning application (opens in a new tab)

  16. Analysis and Implementation of Dense On-Die Memories Using Traditional and Compute-in-Memory Approaches

    For processors and application accelerators executing data-intensive tasks, including machine learning inference, accessing data from dense memory arrays can be a limiting factor for energy efficiency and/or throughput (performance). The high cost of off-chip accesses has led to the extensive use …

    gatech Repository record for Analysis and Implementation of Dense On-Die Memories Using Traditional and Compute-in-Memory Approaches (opens in a new tab)

  17. Benchmark Framework for 2-D/3-D Integrated Compute-in-Memory Based Machine Learning Accelerator

    … frameworks, to evaluate the state-of-the-art compute-in-memory (CIM) accelerators, by considering the hardware performance with the impacts from device options, circuit topologies, architecture hierarchy and data flow, as well as the software accuracy with non-ideal hardware properties. The …

    gatech Repository record for Benchmark Framework for 2-D/3-D Integrated Compute-in-Memory Based Machine Learning Accelerator (opens in a new tab)

Page 1 of 3