惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

M
MIT News - Artificial intelligence
雷峰网
雷峰网
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
Last Week in AI
Last Week in AI
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
阮一峰的网络日志
阮一峰的网络日志
月光博客
月光博客
博客园 - Franky
腾讯CDC
T
Tailwind CSS Blog
Recent Announcements
Recent Announcements
V
V2EX
N
Netflix TechBlog - Medium
量子位
Jina AI
Jina AI
Y
Y Combinator Blog
The GitHub Blog
The GitHub Blog
G
Google Developers Blog
爱范儿
爱范儿
博客园 - 叶小钗
D
Docker
MongoDB | Blog
MongoDB | Blog
D
DataBreaches.Net
T
The Blog of Author Tim Ferriss

cs.DC updates on arXiv.org

DUAL-BLADE: Dual-Path NVMe-Direct KV-Cache Offloading for Edge LLM Inference Progressive Semantic Communication for Efficient Edge-Cloud Vision-Language Models Efficient, VRAM-Constrained xLM Inference on Clients Folding Tensor and Sequence Parallelism for Memory-Efficient Transformer Training & Inference DORA: A Scalable Asynchronous Reinforcement Learning System for Language Model Training AMMA: A Multi-Chiplet Memory-Centric Architecture for Low-Latency 1M Context Attention Serving RaMP: Runtime-Aware Megakernel Polymorphism for Mixture-of-Experts Spark Policy Toolkit: Semantic Contracts and Scalable Execution for Policy Learning in Spark Internet of Everything in the 6G Era: Paradigms, Enablers, Potentials and Future Directions PolyKV: A Shared Asymmetrically-Compressed KV Cache Pool for Multi-Agent LLM Inference A Survey on Split Learning for LLM Fine-Tuning: Models, Systems, and Privacy Optimizations ITAS: A Multi-Agent Architecture for LLM-Based Intelligent Tutoring Latency and Cost of Multi-Agent Intelligent Tutoring at Scale TACO: Efficient Communication Compression of Intermediate Tensors for Scalable Tensor-Parallel LLM Training FreeScale: Distributed Training for Sequence Recommendation Models with Minimal Scaling Cost CommFuse: Hiding Tail Latency via Communication Decomposition and Fusion for Distributed LLM Training A Taxonomy and Resolution Strategy for Client-Level Disagreements in Federated Learning Usable Agent Discovery for Decentralized AI Systems Cloud to Edge: Benchmarking LLM Inference On Hardware-Accelerated Single-Board Computers Data-Free Contribution Estimation in Federated Learning using Gradient von Neumann Entropy Shard the Gradient, Scale the Model: Serverless Federated Aggregation via Gradient Partitioning Promoting Simple Agents: Ensemble Methods for Event-Log Prediction GraphLeap: Decoupling Graph Construction and Convolution for Vision GNN Acceleration on FPGA AGNT2: Autonomous Agent Economies on Interaction-Optimized Layer 2 Infrastructure FedSIR: Spectral Client Identification and Relabeling for Federated Learning with Noisy Labels Stream-CQSA: Avoiding Out-of-Memory in Attention Computation via Flexible Workload Scheduling A Delta-Aware Orchestration Framework for Scalable Multi-Agent Edge Computing Federated Learning over Blockchain-Enabled Cloud Infrastructure Optimal Routing for Federated Learning over Dynamic Satellite Networks: Tractable or Not? Sherpa.ai Privacy-Preserving Multi-Party Entity Alignment without Intersection Disclosure for Noisy Identifiers
Synchronization of ad hoc Clock Networks
Jaderick P. Pabico · 2015-06-25 · via cs.DC updates on arXiv.org

We introduce a graph-theoretic approach to synchronizing clocks in an {\em ad hoc} network of $N$~timepieces. Clocks naturally drift away from being synchronized because of many physical factors. The manual way of clock synchronization suffers from an inherrent propagation of the so called "clock drift" due to "word-of-mouth effect." The current standard way of automated clock synchronization is either via radio band transmission of the global clock or via the software-based Network Time Protocol (NTP). Synchronization via radio band transmission suffers from the wave transmission delay, while the client-server-based NTP does not scale to increased number of clients as well as to unforeseen server overload conditions (e.g., flash crowd and time-of-day effects). Further, the trivial running time of NTP for synchronizing an $N$-node network, where each node is a clock and the NTP server follows a single-port communication model, is~$\bigO(N)$. We introduce in this paper a $\bigO(\log N)$ time for synchronizing the clocks in exchange for an increase of $\bigO(N)$ in space complexity, though through creative "tweaking," we later reduced the space requirement to~$\bigO(1)$. Our graph-theoretic protocol assumes that the network is $\K_N$, while the subset of clocks are in an embedded circulant graph $\C_{n<N}^q$ with $q$~jumps and clock information is communicated through circular shifts within the $\C_{n<N}^q$. All $N$~nodes communicate via a single-port duplex channel model. Theoretically, this synchronization protocol allows for $N(\log N)^{-1} - 1$ more synchronizations than the client-server-based one. Empirically through statistically replicated multi-agent-based microsimulation runs, our protocol allows at most 80\% of the clocks synchronized compared to the current protocol which only allows up to 30\% after some steady-state time.