Search

The expensive cost and small capacity of cache has made prefetching a major method to close the performance gap between main memory and CPU. Prefetching techniques have increasingly shown their importance as a mean to improve cache hit rates. And Metadata is a kind of significant data, especially in the information age, where massive internet productions based on it are created and used in daily life. In this paper, we first narrate the development of prefetching technologies existed to provide a general impression. And two types of prefetching schemes or algorithms, metadata prefetching and adaptive prefetching are introduced from different perspectives. We then compare and analyze their advantages and present some of their limitations. Then, the further discussion shows how to use the technologies depending on different factors in the real situation, which may be the size of data. Besides, the self-learning, as a popular theory is emphasized here for the choice to be combined together with prefetching technology. [ABSTRACT FROM AUTHOR]

Logisim is a digital logic circuit design and simulation software that is an open source, free of charge, secondary development, installation-free, easy to use and intuitive. This paper examines how Logisim can be used to design the data path and combine it with Verilog for FPGA design of single cycle CPUs. [ABSTRACT FROM AUTHOR]

Cloud computing is a computing paradigm that provides various services and computing resources to users through Cloud Service Providers over the Internet. However, storing data in an unsecure cloud may lead to security issues such as privacy issues and data leakage. Therefore, it is necessary for encryption schemes to be implemented in clouds to provide a secure environment for the users. One of the cryptographic schemes is the Attribute Based Encryption (ABE), which provides privacy and access control in cloud and can be implemented in a Trusted Real-Time Execution Environment to achieve stronger security. This paper first outlines the various existing encryption techniques, which can be categorized into symmetric and asymmetric algorithms. It then comprehensively explore the key policy and the access policy attribute based encryption. Furthermore, this paper examines the various schemes of ABE and compares the schemes based on its access structure, advantages and disadvantages. Lastly, this paper discusses the applications of key policy and ciphertext policy ABE. [ABSTRACT FROM AUTHOR]

When conducting a cloud migration in a Trusted Real-Time Execution Environment, it is always important to conduct the migration by following certain standard and constraint while modifying it according to the needs of the migration. Migration framework is introduced to provide such a standard and steps to be followed for migration. However, current migration frameworks no longer essentially satisfy the cloud migration efficiently. Resources are increasing and the usage of current migration frameworks are not proficiently supporting the demands. Thus, researchers are developing more automation cloud migration frameworks that help in reducing the cost, time, manpower and increasing efficiency to conduct cloud migration. This paper concisely addresses cloud migration, stages of conducting cloud migration and introduces various automated cloud migration frameworks along with detailed analysis. [ABSTRACT FROM AUTHOR]

Hadoop Distributed File System (HDFS) is based on Google File System (GFS), a big data distributed file management system included in Hadoop. Nowadays, many HDFS and many other similar frameworks have the need to store small files in the system. In this aspect, HDFS affects its performance and Namenode memory management when dealing with a large number of small files. Therefore, researchers have proposed various solutions to address the shortcomings of HDFS for storing small and medium-sized files. This paper presents three HDFS schemes for merging small files and analyzes the importance of correlation and prefetching after merging small files. The efficiency of reading small files can be improved by correlated file prefetching. Finally, the small file storage architecture is obtained to stand superior to the NHAR architecture. [ABSTRACT FROM AUTHOR]

Nowadays the storing of sensitive information in cyberspace shows a significant increase and this led to an escalation of cyberattacks. Hence, the protection against the confidentiality of data or programs has been a matter of special concern to many security experts. One of the problematic cyberattacks is the cache side-channel attacks as caches are playing an important rule to ensure the performance of computer. The leakage of information can occur by observing the distinct access time for a cache miss or hit. This paper conducts some background research about the categorization of different cache side-channel attacks. Besides that, several strategies which are used to defend against cache side-channel attacks such as detection method, prevention method and some proposed secure cache design will be reviewed. Also, the difficulties about the defenses in cache side-channel attacks and some recommended strategies will also be discussed in the following content. [ABSTRACT FROM AUTHOR]

Deep Neural Networks have gained popularity due to their superiority in performance in various applications. Hence, domain-specific architectures have been introduced to speed up the model training process. Since the launch of TPUs by Google, there were few literatures that examines the performance of TPUs compared to existing processing units such as the CPU and GPU in deep learning applications. The recent inclusion of the PyTorch framework to use Google Colaboratory's resources also provided an opportunity for us to build a PyTorch-based Colab notebook that can be used for further study in this area. In this study, we evaluated the performance of the Intel Xeon CPU, Nvidia T4 GPU and TPU v2-8 by using the Convolutional Neural Network (CNN) model on the CIFAR-10 and MNIST datasets. Our study also considers the accuracy and convergence of the model while testing, which were implemented by data augmentation and batch normalization techniques in the VGG16 model trained from scratch. We achieved a model accuracy of 86% and 96% for the CIFAR-10 and MNIST datasets respectively. During the experiments, we observed that the GPU performed 25x faster than the CPU; meanwhile the TPU only achieved a 12x speedup compared to the CPU as there were some issues faced. We proposed some issues and possible solutions in this paper for further study. [ABSTRACT FROM AUTHOR]

Sky computing, the cloud of clouds, has become an emergent trend in cloud computing domain. It allows the creation of large-scale infrastructure by utilizing resources from multiple cloud providers. It enables these massive infrastructures to run parallel computation with high performance. Being a multi-cloud computing model, sky computing addresses several limitations of traditional cloud computing especially the vendor lock-in problem. Sky computing allows the aggregation of cloud resources among multiple cloud providers so that cloud consumers can dynamically provision the required resource on demand. This paper reviews state-of-the-art and explains their significance in the development of a sky computing environment. [ABSTRACT FROM AUTHOR]