Results 191 to 200 of about 9,496 (247)

Boosting Write Performance of KV Stores: An NVM - Enabled Storage Collaboration Approach

IEEE International Conference on Data Engineering
As the most common data structure for key-value stores, LogStructured Merge Tree (LSM-tree) can eliminate random write operations and keep acceptable read performance.
Yi Wang, Jia-Jian He, Kaoyi Sun, Yun-Hao Dong, Jia-Xian Chen, Chen-Lin Ma, Amelie Chi Zhou, Rui Mao   +7 more
semanticscholar   +1 more source

Novel Weight Mapping Method for Reliable NVM based Neural Network

IEEE International Reliability Physics Symposium, 2021
Errors of vector-matrix-multiplication induced by interconnect resistance become a crucial reliability challenge in non-volatile memory (NVM) based neural network. Here, we propose a novel weight mapping method, called weight mapping correction (WMC), to
Li-Xia Han, Ya-Chen Xiang, P. Huang, Gui-Hai Yu, Runze Han, Xiao-Yan Liu, Jin-Feng Kang   +6 more
semanticscholar   +1 more source

NVM-Enhanced Machine Learning Inference in 6G Edge Computing

IEEE Transactions on Network Science and Engineering
With the increasing popularization of smart terminals and real-time interactive applications, fast growing technical requirements push both academia and industry to look beyond 5G and conceptualize the sixth generation (6G) mobile network.
Xiao-Jun Shang, Yaodong Huang, Zhen-Hua Liu, Yuan-Yuan Yang   +3 more
semanticscholar   +1 more source

HiLSM: an LSM-based key-value store for hybrid NVM-SSD storage systems

ACM International Conference on Computing Frontiers, 2020
In order to ensure data durability and crash consistency, the LSM-tree based key-value stores suffer from high WAL synchronization overhead. Fortunately, the advent of NVM offers an opportunity to address this issue.
Wenjie Li, De-Jun Jiang, Jin Xiong, Y. Bao   +3 more
semanticscholar   +1 more source

Enabling Secure NVM-Based in-Memory Neural Network Computing by Sparse Fast Gradient Encryption

IEEE transactions on computers, 2020
Neural network (NN) computing is energy-consuming on traditional computing systems, owing to the inherent memory wall bottleneck of the von Neumann architecture and the Moore's Law being approaching the end.
Yi Cai, Xiao-Ming Chen, Lu Tian, Yu Wang, Hua-Zhong Yang   +4 more
semanticscholar   +1 more source

Unbounded Hardware Transactional Memory for a Hybrid DRAM/NVM Memory System

Micro, 2020
Persistent memory programming requires failure atomicity. To achieve this in an efficient manner, recent proposals use hardware-based logging for atomic-durable updates and hardware transactional memory (HTM) for isolation.
Jungi Jeong, Jaewan Hong, S. Maeng, Changhee Jung, Youngjin Kwon   +4 more
semanticscholar   +1 more source

The inevitable rise of NVM in computing

2010 IEEE Hot Chips 22 Symposium (HCS), 2010
This article consists of a collection of slides from the author's conference presentation of a seminar discussing the development and market demand for nonvolatile memory (NVM). Some of the specific topics discussed include: market and demand analyses for NVM technologies; systems that support NVM and applications supported; how flash memories will ...
openaire   +2 more sources

Fine-grained metadata journaling on NVM

2016 32nd Symposium on Mass Storage Systems and Technologies (MSST), 2016
Journaling file systems have been widely used where data consistency must be assured. However, we observed that the overhead of journaling can cause up to 48.2% performance drop under certain kinds of workloads. On the other hand, the emerging high-performance, byte-addressable Non-volatile Memory (NVM) has the potential to minimize such overhead by ...
Cheng Chen 0008, Jun Yang 0022, Qingsong Wei, Chundong Wang 0001, Mingdi Xue   +4 more
openaire   +2 more sources

Fast journaling made simple with NVM

Proceedings of the 35th Annual ACM Symposium on Applied Computing, 2020
Journaling file systems (JFS), which base their architecture on traditional database logging, write journals sequentially through a form of centralized logging, primarily to restore file systems in the face of an unexpected failure. Due to the extra write operations required for journaling, the performance of the JFS is negatively impacted, thus ...
Woong Sul, Kihwang Kim, Minsoo Ryu, Hyungsoo Jung 0001, Hyuck Han   +4 more
openaire   +2 more sources

Safe and Efficient Remote Application Code Execution on Disaggregated NVM Storage with eBPF

arXiv.org, 2020
With rapid improvements in NVM storage devices, the performance bottleneck is gradually shifting to the network, thus giving rise to the notion of "data movement wall".
K. Kourtis, A. Trivedi, Nikolas Ioannou
semanticscholar   +1 more source