Aug 2026· Asia-Pacific Workshop on Networking· pp. 315-317· 0 citations· 16 references
Computer Science
TL;DR
Quanta is presented, a redundancy-aware simulation framework that eliminates repeated execution in packet-level simulation and reduces the dependence of simulation cost on physical network scale.
Abstract
Packet-level network simulation provides high-fidelity modeling but suffers from severe scalability bottlenecks. Existing scaling approaches remain inefficient for modern data-center and AI-training networks. Spatial parallelism requires substantial hardware resources, while temporal-skipping approaches become less effective under bursty traffic. We observe that homogeneous data-center deployments introduce substantial execution redundancy during simulation. This paper presents Quanta, a redundancy-aware simulation framework that eliminates repeated execution in packet-level simulation. Quanta reduces the dependence of simulation cost on physical network scale. Our evaluation shows that Quanta accelerates large-scale simulations by up to 90 × , synergizes with parallel execution for a 320 × combined speedup.
LST-Sim is presented, an efficient simulator designed for large-scale model training, balancing simulation accuracy and runtime efficiency, and a domain-based traffic management mechanism that integrates theoretical analysis with network simulation to handle different types of communication traffic, thereby improving simulation efficiency.
Siwei Ji, Yihao Sun, Bo Lei et al.· Proceedings of the ACM SIGCO...· 0 citations
In the performance evaluation of communication networks, packet-level simulation entails enormous computational costs under conditions such as large-scale networks or numerous communicating flows, leading to issues where execution time exceeds practical limits. In contrast, flow-level simulation, which approximates traffic as a continuous fluid, can mitigate the increase in computational complexity associated with the growing number of packets. We have previously proposed a method to accelerate simulations by circulating a token, which carries information regarding the network state, along the path traversed by a flow during the simulation. However, existing token-based simulators assume a fixed step size and perform state updates at constant intervals, even during periods where state changes are minimal. Consequently, redundant computations occur relative to the required accuracy. In this study, we propose a method that introduces dynamic step size control to a token-based flow-level simulator. To address the structural issue of token history omissions caused by step size alterations, we propose a hybrid step-doubling method that estimates missing token information through token interpolation and direct calculation of internal states, thereby autonomously controlling the step size. Evaluation experiments conducted in a simple environment using a dumbbell network topology with a single bottleneck confirmed the effectiveness of the proposed method. Compared to the conventional fixed step size method, the simulation execution time was reduced by approximately 90% to 97%, and memory consumption was reduced by approximately 25% to 70%.
To overcome data movement bottlenecks, we propose a unified host and device interaction architecture for computational storage. Our framework facilitates data offloading through a host dispatch engine and a device admission controller. Evaluated using a storage simulator across vector and genomic datasets, our prototype demonstrates substantial speedups for tasks with high data reduction ratios. However, evaluation reveals performance degradation in scenarios with heavy return data, highlighting the critical limitations of static offloading. These findings validate the necessity of developing dynamic offloading strategies for future architectures.
Wei-Chun Cheng, Ti-Wei Tsai, H. Wei et al.· IEEE Non-Volatile Memory Sys...· 0 citations
Evaluation results show that, while these mechanisms consume a small amount of link bandwidth, CBFC can greatly reduce receive buffer utilization, and LLR can substantially mitigate network performance degradation caused by packet corruption.
Kefei Liu, Ruixue Wang, Tianrun Jiang et al.· Proceedings of the ACM SIGCO...· 0 citations
The proposed Multi-Path Multi-Level Feedback Queueing (MP-MLFQ) leverages the spatial diversity and regularity of DCNs to realize a scheduler with numerous logical priority levels while occupying as low as 2 physical priority queues within network switches.
Alessandro Cornacchia, Andrea Bianco, Paolo Giaccone et al.· 0 citations
With the rapid growth of large language model training and generative artificial intelligence services, data center networks face severe micro-burst traffic and high concurrency. Traditional hash-based flow-level load balancing cannot sense link states, leading to hash collisions, hotspot congestion, and tail latency in multipath Clos networks. Existing packet-level schemes are constrained by stale state information, high hardware complexity, and poor adaptation to heterogeneous links. To address these issues, this paper proposes probabilistic state-proportional (PSP) dispatching, a packet-level load balancing algorithm. Using a Band-based discrete state representation, PSP replaces global sorting with local probability mapping, reducing hardware complexity while suppressing herding and oscillations caused by stale states. Experiments on a cycle-accurate simulator show that PSP is robust across port scales, bandwidth-limited paths, and fixed-flow interference. It outperforms join-the-shortest-queue (JSQ) scheduling and Random in loss rate, 99th-percentile buffer occupancy, and scalability, while remaining competitive with Top-k at lower hardware cost. PSP provides an effective balance among performance, stability, and overhead for artificial intelligence data centers.
Jiaqi Liu, Chunyang Zhang, Heng Pan et al.· 0 citations