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

博客园 - 【当耐特】
N
Netflix TechBlog - Medium
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
雷峰网
雷峰网
MongoDB | Blog
MongoDB | Blog
有赞技术团队
有赞技术团队
Engineering at Meta
Engineering at Meta
M
MIT News - Artificial intelligence
Google DeepMind News
Google DeepMind News
罗磊的独立博客
Hugging Face - Blog
Hugging Face - Blog
WordPress大学
WordPress大学
T
Tailwind CSS Blog
小众软件
小众软件
J
Java Code Geeks
人人都是产品经理
人人都是产品经理
博客园_首页
MyScale Blog
MyScale Blog
博客园 - 聂微东
V
Visual Studio Blog
The Cloudflare Blog
月光博客
月光博客
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
U
Unit 42

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
High Performance Consensus without Duplication: Multi-pip...
Taining Cheng · 2022-05-09 · via cs.DC updates on arXiv.org

The state-of-the-art HotStuff operates an efficient pipeline in which a stable leader drives decisions with linear communication and two round-trips of message. However, the unifying proposing-voting pattern is not sufficient to improve the bandwidth and concurrency performance of the modern system. In addition, the delay corresponding to two rounds of message to produce a certified proposal in that scheme is a significant performance bottleneck. Thus, this study developed a new consensus protocol, Multi-pipeline HotStuff, for permissioned blockchain. To the best of the authors' knowledge, this is the first protocol that combines multiple HotStuff instances to propose batches in order without a concurrent proposal, such that proposals are made optimistically when a correct replica realizes that the current proposal is valid and will be certified by quorum votes in the near future. Because simultaneous proposing and voting are allowed by the proposed protocol without transaction duplication, it produced more proposals in every two rounds of messages. In addition, it further boosted the throughput at a comparable latency with that of HotStuff. The evaluation experiment conducted confirmed that the throughput of Multi-pipeline HotStuff outperformed that of the state-of-the-art protocols by approximately 60\% without significantly increasing end-to-end latency under varying system sizes. Moreover, the proposed optimization also performed better when it suffers a bad network condition.