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 “"Inference Acceleration"”.
-
MEMORY-CENTRIC ARCHITECTURES FOR AI INFERENCE ACCELERATION VIA HARDWARE–SOFTWARE CO-DESIGN
… wall. This thesis advances memory-centric AI inference through cross-layer hardware-software co-design spanning analog and digital in-memory computing (IMC), near-memory computing (NMC), and communication-aware system architectures. It first analyzes device- and circuit-level non-idealities in …
-
Toward Efficient Deep Models : From Weight Optimization and Inference Acceleration to Prompt Optimization
Thesis (Ph.D.)--Michigan State University. Computer Science - Doctor of Philosophy, 2026
-
Silicon photonic hardware accelerators for transformers and graph neural networks
… imposes unique challenges, hindering their inference and usage in resource-constrained systems. For instance, the transformer model's size, number of parameters, and complexity of operations lead to long inference times, large memory footprint, and low computation-to-memory ratio. On the …
-
Toward Efficient, Scalable, And Reliable Edge-Enabled Applications For Connected Vehicles
… the key advantages of bandwidth reduction and inference acceleration for machine learning models. Then, a collaborative learning framework based on a group of heterogeneous computing platforms will be presented, which enables collaborative training (i.e., electric vehicle battery failure …
-
Architecture of multiple processing units for accelerating deep neural network inference
Deep neural network (DNN) inference increasingly relies on specialized hardware for high computational efficiency. This thesis presents a scalable architecture for DNN inference acceleration, prototyped on an FPGA, with a focus on general matrix-matrix multiplication (GEMM)-based operations. Also, …
-
Architecture and Circuit Design Optimization for Compute-In-Memory
… (ADC) introduce significant overhead in CIM inference design, the research first focuses on the circuit optimization for inference acceleration and proposes a resistive random access memory (RRAM) based ADC-free in-memory compute scheme. We comprehensively explore the trade-offs involving …
-
COMPUTE-IN-MEMORY WITH EMERGING NON-VOLATILE MEMORIES FOR ACCELERATING DEEP NEURAL NETWORKS
… 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 which monolithically integrate the RRAM memory array and CMOS peripheral circuits are designed and fabricated using …
-
Computer-aided design of atomic silicon quantum dots and computational applications
… application level. This work identifies hardware acceleration of machine learning inference as a novel prospective application on the DB computational platform. Matrix multiplication is a common operation in the inference stage of many neural network implementations, and presents a bottleneck to …
-
Alternating Optimization: Constrained Problems, Adversarial Networks, and Robust Models
… training a model from large-scale samples, and inference on new samples after the model is deployed. The training of modern models relies on solving difficult optimization problems that involve nonconvex, nondifferentiable objective functions and constraints, which is sometimes slow and often …
-
Reconfigurable Architectures for Deep Learning
… making them suitable for accelerating DL inference. In the first part of this thesis, we evaluate the security and performance of current FPGAs for DL inference acceleration. We demonstrate that the inherent resilience of DL models to computation mistakes can provide additional protection …
-
High performance deep learning on resource constrained platforms
… these from leaving the Cloud, where model inference takes place. As a result, only a fraction of ML-powered applications, such as keyword spotting and next-word prediction, remain lightweight enough in terms of memory and compute footprint to run on-device.</p> <p>Motivated by the urge to …
-
IMPROVING MODEL AND DATA EFFICIENCY FOR DEEP LEARNING
… have been made to accelerate the training and inference by methods such as quantization and pruning. Regarding data efficiency, few-shot learning, semi-supervised learning, and self-supervised learning have gathered more attention due to their abilities to learn feature representations with few …
-
Memory-aware fusing and tiling of neural networks for accelerated edge inference
A rising research challenge is running costly machine learning (ML) networks locally on resource-constrained edge devices. ML networks with large convolutional layers can easily exceed available memory, increasing latency due to excessive swapping. Previous memory reduction techniques such as …
-
Language Model Inference on FPGA with Integer-only Operations
… this challenge by investigating the integer-only acceleration of transformer models on FPGAs, focusing on the BERT architecture. We demonstrate that by removing the floating-point operations from the inference pipeline, especially from non-linear functions like GELU, Softmax, and Layer …
-
Accelerating Token Generation in Large Reasoning Models Using FlashHead
… reasoning traces, but this increases inference la tency under autoregressive decoding. A significant part of this cost can come from the language modeling head, which computes logits over the full vocabulary at ev ery decoding step. This thesis studies FlashHead, a training-free …
-
Smarter, Leaner, Faster: Practical Efficiency for Democratizing Large Language Models
… on similar layer importance metrics, cutting inference compute and the memory requirement retaining performance. These techniques work to realign the cost curve of large language models. By converting redundancy into cheap tokens, allocating token budgets to only the problems that matter, and …
-
Hardware-software codesign of silicon photonic AI accelerators
… throughput, low latency, and energy-efficient AI acceleration across various neural network models, using silicon photonics. At the architectural level, we have proposed wavelength reuse schemes, vector decomposition, and NN-aware MAC unit designs for increased efficiency in laser power …
-
Skip, Search, Speculate - Turning a Frozen LLM into Its Own Lightweight Drafter for Lossless Inference Speedup
… has investigated the possibility to speedup inference by producing a self-speculative setup where the original model is verifier and drafter. The model is frozen so the original quality is preserved and no retraining required. The frozen model switches between being verifier and drafter. In …
-
Συνελικτικά vευρωνικά δίκτυα σε υπολογιστικές διατάξεις τεχνολογίας IMC (In-Memory Computing)
… with strict resource constraints. The efficient acceleration of the inference process in such environments requires architectures that drastically reduce the cost of accessing and transferring data to and from memory, addressing the so-called memory wall, which is exacerbated by the large volume …