scispace - formally typeset
Search or ask a question

Answers from top 9 papers

More filters
Papers (9)Insight
Proceedings ArticleDOI
13 Jun 2013
4 Citations
A virtual server that has an overload can get memory resources from another virtual server that has less workload and it has memory unused.
Open accessProceedings ArticleDOI
Daniele Sciascia, Fernando Pedone 
08 Oct 2012
9 Citations
RAM-DUR's key insight is a sophisticated distributed cache mechanism that provides high performance and strong consistency without the limitations of existing solutions (e. g., no single server must have enough memory to cache the entire database).
Our experiments show that the optimal buffer allocation shifts to placing more memory at the server as the server has progressively less information about future frame sizes.
This structure requires much less RAM.
Book ChapterDOI
S. Dov Gordon, Jonathan Katz, Xiao Wang 
02 Dec 2018
24 Citations
We show a protocol for two-server oblivious RAM (ORAM) that is simpler and more efficient than the best prior work.
Phase-change RAM, magnetic RAM, and resistive RAM offer strong scalability, speed, and power consumption advantages over conventional capacitance-based memory.
Open accessBook ChapterDOI
Steve Lu, Rafail Ostrovsky 
03 Mar 2013
155 Citations
As alluded above, our two-server Oblivious RAM protocol leads to a novel application in the realm of secure two-party RAM program computation.
Proceedings ArticleDOI
Daesung Lee, Kuinam J. Kim 
21 Apr 2010
7 Citations
But the use of a cache server can cause another bottleneck because of the concentration of requests at the cache server.
In this paper we present two new algorithms which significantly enhance the ability of Q-RAM to make resource tradeoff decisions.

See what other people are reading

What are the problems with internet of things?
5 answers
The Internet of Things (IoT) faces several challenges, including security and privacy issues. These problems stem from factors such as inadequate device updates, user inexperience, lack of strong security measures, and unreliable device tracking. Researchers have highlighted concerns like outdated components in devices, disclosure of sensitive information, and devices configured with insecure default settings. To address these issues, proposed solutions include implementing security agreements, developing layered security models, and utilizing SSL certificates for secure data transfer between IoT devices. The vulnerabilities identified in IoT systems, such as webcams with outdated components and devices with default passwords, underscore the urgent need for robust security measures and standardized protocols in IoT environments.
How do cashing techniques impact the optimization of models for edge deployment?
5 answers
Caching techniques play a crucial role in optimizing models for edge deployment by reducing computational complexity and storage requirements. The use of caching helps in maximizing the efficiency of AI models by minimizing the resources required for real-time processing on edge devices. Additionally, caching aids in reducing the computational expense for edge devices by optimizing the memory usage and computational speed of neural networks deployed at the edge. Furthermore, caching techniques can enhance the end-to-end latency and reduce monetary costs significantly when compared to traditional cloud deployment, making them essential for efficient edge computing. In the context of Edge Cloud Computing, caching strategies can contribute to more efficient resource utilization and high-quality service delivery by carefully selecting deployment locations based on machine learning algorithms.
How does the use of Excel impact productivity and efficiency in the workplace?
5 answers
The use of Excel in the workplace significantly impacts productivity and efficiency. Research indicates that Excel is a dominant spreadsheet editor widely used by organizations. Despite its widespread use, many users only utilize relatively low-level features, highlighting the need for more comprehensive training to fully capitalize on Excel's potential to enhance workplace performance. Studies also suggest that end-user spreadsheets often contain errors, prompting the proposal of a model-driven engineering approach to improve productivity. This approach aims to enhance effectiveness and efficiency by providing users with a structured model-based environment for spreadsheet tasks, potentially reducing errors and increasing productivity. Overall, Excel's impact on productivity is substantial, but there is room for improvement through better training and innovative approaches to spreadsheet usage.
Is google cloud vision accurate for character recognition for english alphabets and digits?
5 answers
Google Cloud Vision demonstrates high accuracy in character recognition for English alphabets and digits. In a study comparing OCR performance, Google Cloud Vision API achieved an accuracy of 84.43% for recognizing characters from a Thai vehicle registration certificate. Additionally, the system utilizing Google Cloud Vision for OCR tasks on Colombian vehicle license plates showed a 100% accuracy rate, highlighting its effectiveness. These results indicate that Google Cloud Vision is reliable for character recognition tasks. The use of cloud-based neural networks like CNN for character recognition further enhances accuracy and computational efficiency. Therefore, Google Cloud Vision is a suitable choice for accurate character recognition of English alphabets and digits, offering high precision and readability.
What scalability method makes possible the distribution of data across partitions in NoSQL (MongoDB) databases?
5 answers
Database partitioning is a crucial scalability method in NoSQL databases like MongoDB, enabling the distribution of data across partitions for enhanced performance. In MongoDB, horizontal partitioning is achieved through sharding, where each shard functions as a replica set, ensuring data distribution and load balancing. Automated table splitting mechanisms that consider workload patterns are proposed to improve load balancing in large-scale scenarios, addressing the limitations of manual partitioning based solely on table size. Additionally, efficient algorithms like GPA are utilized to group denormalized columns into families, optimizing query processing performance during data conversion from SQL to NoSQL databases. Furthermore, advanced systems like Grep leverage graph models and neural networks to automate database partitioning based on data and query correlations, significantly enhancing performance in distributed databases.
How do environmental filters differ between ecoclimates?
5 answers
Environmental filters vary between ecoclimates due to factors like climate harshness, paleoclimatic changes, and dispersal limitations. In colder and drier environments, abiotic filtering intensifies, impacting species distributions significantly. Climate niche values play a crucial role in filtering vegetation composition, with frost having a strong average effect. Environmental filtering, especially climatic filtering, enhances model performance in ecological niche modeling, emphasizing the importance of reducing sampling biases. Climate characteristics act as key filters for organisms, affecting community compositions, dynamics, and taxonomic relatedness patterns. In transcontinental naturalization, abiotic filters like climate conditions play a significant role in determining species richness, especially the presence of exotic species in different regions. Climatic and geographic filtering hypotheses influence largescale diversity patterns of tree species communities, showcasing the impact of climatic harshness and paleoclimatic changes on species sorting and diversity patterns.
Why is effective task management in team work important for team performance?
5 answers
Effective task management in team work is crucial for enhancing team performance. Research has shown that managing tasks efficiently leads to improved collaboration and productivity. By utilizing innovative systems like bots within communication channels, teams can streamline task creation, assignment, and tracking, reducing friction and inefficiencies. Additionally, structured group dynamics and tailored team formations contribute to providing high-quality services, fostering better relationships at work, and ultimately achieving greater organizational success. Furthermore, optimizing task allocation through methods like transferring tasks to servers with an affinity for them enhances overall task efficiency and performance. Overall, effective task management ensures smoother workflows, better coordination, and ultimately boosts team productivity and success.
How IoT devices can be cost-effective for water utilities, especially when budgets are tight and technology upgrades are costly?
5 answers
IoT devices offer cost-effective solutions for water utilities, particularly beneficial when facing budget constraints and expensive technology upgrades. These devices enable real-time monitoring of water quality and quantity, aiding in efficient water distribution, preventing wastage, and ensuring safe water supply. By utilizing sensors and cloud-based systems, IoT devices provide continuous data collection and analysis, facilitating proactive maintenance and resource management. Additionally, IoT technology can detect anomalies like water leaks and faults in pipes, reducing expenses on water bills and preventing structural damage. Overall, the integration of IoT devices in water utilities optimizes operations, enhances cost-effectiveness, and promotes sustainable water management practices within limited budgets.
What blockchain can and can’t do: Applications to marketing and privacy?
5 answers
Blockchain technology offers distributed solutions for various applications like finance, healthcare, and cybersecurity. While it ensures transparency and reliability in transaction ledgers, it faces challenges in maintaining user privacy due to the public nature of data on the blockchain. To address privacy concerns, solutions such as secure multiparty computation protocols and private blockchains have been proposed. However, the trade-off between security and privacy remains a critical consideration when implementing blockchain in different sectors. In marketing, blockchain can revolutionize processes by providing secure and transparent transactions. Nevertheless, the exposure of sensitive data on a public blockchain poses privacy risks that need to be carefully managed to fully leverage the benefits of blockchain technology.
What is power consumption per server in data centers?
5 answers
The power consumption per server in data centers varies based on different models and methodologies. Various studies have proposed models to estimate server power consumption accurately. One approach involves using a cubic polynomial IT equipment model (CPIT) to represent server power consumption, showing superior precision compared to existing models. Another study suggests utilizing Artificial Neural Networks (ANN) to predict server power consumption, demonstrating better performance than traditional mathematical formulas. Additionally, a model combining Principal Component Analysis (PCA) and DeepAR network has been proposed to predict server power consumption, outperforming other models in terms of accuracy. Furthermore, a novel power monitoring approach using electromagnetic interference for power estimation shows promising results with a mean absolute percentage error of ~3%.
How does the placement of edge servers affect overall latency in a network?
5 answers
The placement of edge servers has a significant impact on overall latency in a network. Efficient placement techniques aim to select optimal locations for edge servers to improve network performance. By strategically placing edge servers, the balance between energy consumption and latency can be achieved. Several factors need to be considered, such as user resource demands, user mobility, and network traffic distribution. Techniques like particle swarm optimization (PSO) and heuristic algorithms have been proposed to optimize network delay and energy consumption, especially in high mobility areas. Experimental results have shown that these approaches can reduce edge servers' total cost and overloaded numbers, leading to improved network performance.