Non-volatile memory (NVM) devices, such as Flash, phase change RAM, spin transfer torque and resistive offer several advantages challenges when compared to conventional technologies, DRAM magnetic hard disk drives (HDDs). In this paper, we present a survey of software techniques that have been proposed exploit the mitigate disadvantages NVMs used for designing systems, and, in particular, secondary storage (e.g., solid state drive) main memory. We classify these along dimensions highlight...

In recent years, there has been an enormous interest in using deep learning to classify underwater images identify various objects, such as fishes, plankton, coral reefs, seagrass, submarines, and gestures of sea divers. This classification is essential for measuring the water bodies' health quality protecting endangered species. Furthermore, it applications oceanography, marine economy defense, environment protection, exploration, human-robot collaborative tasks. article presents a survey...

Recent technological advances have greatly improved the performance and features of embedded systems.With number just mobile devices now reaching nearly equal to population earth, systems truly become ubiquitous.These trends, however, also made task managing their power consumption extremely challenging.In recent years, several techniques been proposed address this issue.In paper, we survey for systems.We discuss need management provide a classification on important parameters highlight...

As data movement operations and power-budget become key bottlenecks in the design of computing systems, interest unconventional approaches such as processing-in-memory (PIM), machine learning (ML), especially neural network (NN)-based accelerators has grown significantly. Resistive random access memory (ReRAM) is a promising technology for efficiently architecting PIM- NN-based due to its capabilities work both: High-density/low-energy storage in-memory computation/search engine. In this...

Recent trends of CMOS scaling and increasing number on-chip cores have led to a large increase in the size caches. Since SRAM has low density consumes amount leakage power, its use designing caches become more challenging. To address this issue, researchers are exploring several emerging memory technologies, such as embedded DRAM, spin transfer torque RAM, resistive phase change RAM domain wall memory. In paper, we survey architectural approaches proposed for systems and, specifically, with...

The continuous drive for performance has pushed the researchers to explore novel memory technologies (e.g. nonvolatile memory) and fabrication approaches 3D stacking) in design of caches. However, a comprehensive tool which models both conventional emerging 2D designs been lacking. We present DESTINY, microarchitecture-level modeling (and 2D) cache using SRAM, embedded DRAM (eDRAM), spin transfer torque RAM (STT-RAM), resistive (ReRAM) phase change (PCM). DESTINY facilitates design-space...

For extreme-scale high-performance computing systems, system-wide power consumption has been identified as one of the key constraints moving forward, where DRAM main memory systems account for about 30 to 50 percent a node's overall consumption. As benefits device scaling slow, it will become increasingly difficult keep capacities balanced with increasing computational rates offered by next-generation processors. However, several emerging technologies related nonvolatile (NVM) devices are...

The continuous drive for performance has pushed the researchers to explore novel memory technologies (e.g. non-volatile memory) and fabrication approaches 3D stacking) in design of caches. However, a comprehensive tool which models both conventional emerging 2D designs been lacking. We present DESTINY, microarchitecture-level modeling (and 2D) cache using SRAM, embedded DRAM (eDRAM), spin transfer torque RAM (STT-RAM), resistive (ReRAM) phase change (PCM). DESTINY facilitates design-space...

As the number of cores on a chip increases and key applications become even more data-intensive, memory systems in modern processors have to deal with increasingly large amount data. In face such challenges, data compression presents as promising approach increase effective system capacity also provide performance energy advantages. This paper survey techniques for using cache main systems. It classifies based parameters highlight their similarities differences. discusses CPUs GPUs,...

During the last decade, there has been considerable interest of researchers towards use two-dimensional (2D) materials for electronic device implementations. The main driving force is improved performance offered by these 2D operation in nano-scale regime. Among material, silicene (the silicon) emerged as preferred choice because its expected integration with silicon based technology. This technology one primary advantages a material future devices availability infrastructure bulk...

The capability of the self-attention mechanism to model long-range dependencies has catapulted its deployment in vision models. Unlike convolution operators, offers infinite receptive field and enables compute-efficient modeling global dependencies. However, existing state-of-the-art attention mechanisms incur high compute and/or parameter overheads, hence unfit for compact convolutional neural networks (CNNs). In this work, we propose a simple yet effective "Ultra-Lightweight Subspace...

In recent years, continuous latent space (CLS) and discrete (DLS) deep learning models have been proposed for medical image analysis improved performance. However, these encounter distinct challenges. CLS capture intricate details but often lack interpretability in terms of structural representation robustness due to their emphasis on low-level features. Conversely, DLS offer interpretability, robustness, the ability coarse-grained information thanks structured space. limited efficacy...

Recent trends of aggressive technology scaling have greatly exacerbated the occurrences and impact faults in computing systems. This has made `reliability' a first-order design constraint. To address challenges reliability, several techniques been proposed. paper provides survey architectural for improving resilience We especially focus on proposed microarchitectural components, such as processor registers, functional units, cache main memory etc. In addition, we discuss non-volatile memory,...