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 27 for “"Software/Hardware Co-Design"”.
-
Software-hardware co-design for energy efficient datacenter computing
Datacenters have become commonplace computing environments used to offload applications from distributed local machines to centralized environments. Datacenters offer increased performance and efficiency, reliability and security guarantees, and reduced costs relative to independently operating the …
-
Software-Hardware Co-design For Deep Learning Model Acceleration
… sizes and numbers of operations. However, it becomes a great challenge to deploy modern DNNs on mobile and edge devices because of their limited memory, computation resources and battery energy. This dissertation seeks to address these challenges by advancing and integrating techniques from both …
-
Analytically driven software/hardware co-design for accelerating tensor workloads
… is a difficult problem due to the enormous, non-convex, and non-differentiable search space. As workloads become larger and larger, the problem becomes harder while the importance of dataflow increases. To tackle the problem, prior work reduces the search space using empirically driven, or …
-
Software-Hardware Co-Design of Efficient Rendering and Processing for Modern AR/VR
… to longer frame times and potential motion discomfort. Techniques like post-processing aim to enhance detail but can strain GPU capabilities. In contrast, AR overlays digital content onto the real world, enabling real-time interaction but facing issues like lengthy initialization times and high …
-
Resource-Efficient Software-Hardware Co-Design for Voxelized 3D Point-Cloud’s Indexing and Storage
… systems, specifically with the aid of a software-hardware computer co-design for the indexing and storage of voxelized 3D point-cloud streams, has the potential to prolong those systems' uptime. 3D maps, with many millions to billions of points, are now constructed and used in an …
-
On the importance of software-hardware co-design of Trusted Computing architectures and services
The increasing heterogeneity and connectivity of computer systems, from low-end embedded devices to high-end servers, present a unique security challenge. The ubiquity of these systems and their role in critical infrastructure has made them prime targets for cyberattacks, with potentially …
-
Software/Hardware Co-Design to Improve Productivity, Portability, and Performance of Loop-Task Parallel Applications
Computer architects are increasingly turning to programmable accelerators tailored for narrower classes of applications in order to achieve high performance and energy efficiency. A continuing challenge with accelerators is enabling the programmer to easily extract maximum performance without …
-
Algorithm Optimization and Hardware Acceleration for Machine Learning Applications on Low-energy Systems
… triumph in its application field, the increasing complexity of the learning models introduces neoteric challenges to the ML system designs. On the one hand, the applications of ML on resource-restricted terminals, like mobile computing and IoT devices, are prevented by the high computational …
-
Συν-σχεδιασμός υλικού / λογισμικού ενσωματωμένων συστημάτων χαμηλής κατανάλωσης ισχύος για εφαρμογές επεξεργασίας σήματος
… introduced into the market, having managed to become an essential part of our daily lives. The design and implementation of embedded systems is still an open issue posing many challenges to researchers all over the world. The power consumption is an important constraint and must be taken into …
-
Digital Signal Processors and physical implementation for large scale microelectrode array local field potential signal segmentation
… the population and 30% of patients resistant to conventional therapies. This challenge has drawn increasing attention from engineering and neuroscience, driving the need for implantable brain-based solutions to improve treatment efficacy and patient outcomes. Neural implants have become crucial …
-
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 …
-
Secure and Energy-Efficient High-Performance Computing
High-performance computing (HPC) is on the verge of the exascale era, yet continued scaling is impeded by three convergent threats: endemic memory-safety constraints in the dominant distributed-memory model and runtime systems, micro-architectural side channels that erode isolation on heterogeneous …
-
A Hybrid Paradigm Compilation System for Efficient SNN Inference on the Heterogeneous Multi-core SpiNNaker2 Neuromorphic Platform
… have the potential to deliver energy-efficient computation, but their performance strongly depends on how well the underlying hardware is utilized. Modern neuromorphic systems, such as SpiNNaker2, integrate both serial processors and parallel accelerators, offering new opportunities for …
-
Multi-resource Fair Scheduler in Linux
… of operating systems. In Linux, the Completely Fair Scheduler (CFS) allocates CPU time to processes, allowing them to share CPU time. Although effective in allocating CPU time, CFS does not consider the consumption of other system resources (e.g., memory bandwidth, I/O devices, and …
-
Tapered-Precision Numerical Formats for Deep Learning Inference and Training
… with current deep learning models, is their computational complexity. One approach to address this limitation is to compress the deep learning models by employing low-precision numerical formats. Such low-precision models often suffer from degraded inference or training accuracy. This lends …
-
Methodologies for co-designing supercomputer-scale systems and deep learning software
This dissertation introduces new methodologies to co-design deep learning software and supercomputer hardware in the setting of large-scale training. The first is an analytical performance model for exploring the co-design space of parallel algorithms for large language models (LLMs) and potential …
-
Language Support for Real-time Data Processing
… (1) creating a language for specifying complex computations over real-time data streams, (2) developing software-hardware co-design to efficiently match regular patterns in a streaming setting, and (3) designing a system for parallel stream processing with the preservation of sequential …
-
Reactive processing for synchronous languages and its worst case reaction time analysis
… systems. These are systems that have to react continuously to the environment at a rate that is determined by the environment. Reactive systems have two specific characteristics : their control flow requires concurrency and preemption, and, since the reactive systems are often safety-critical, …
-
Toward Fast Deployment and Resource Efficient AI on FPGAs
As machine learning models flourish, many hardware accelerators have been proposed, but designing an application specific accelerator for each new model or algorithm is not straightforward and usually requires multiple iterations between hardware and software. Field programmable gate arrays (FPGAs) …
-
TACKLING PERFORMANCE AND SECURITY ISSUES FOR CLOUD STORAGE SYSTEMS
Building data-intensive applications and emerging computing paradigm (e.g., Machine Learning (ML), Artificial Intelligence (AI), Internet of Things (IoT) in cloud computing environments is becoming a norm, given the many advantages in scalability, reliability, security and performance. However, …
Page 1 of 2