Performance Evaluation最新文献

{"title":"ASIP tandem queues with Lévy input and consumption","authors":"Onno Boxma ,&nbsp;Offer Kella ,&nbsp;Jacques Resing","doi":"10.1016/j.peva.2025.102513","DOIUrl":"10.1016/j.peva.2025.102513","url":null,"abstract":"<div><div>We consider an ASIP (asymmetric inclusion process) tandem queue, in which the first queue receives a fluid input according to a nondecreasing Lévy process. Each queue has a gate that opens after independent, exponentially distributed periods for an infinitesimal amount of time, allowing the queue content to move to the next queue. In addition, again at independent exponentially distributed instants, a fixed fraction of a queue content is removed from the system.</div><div>For this model, restricting ourselves to steady state, we obtain the following results. (i) We derive the buffer content distribution of the first queue. (ii) For the 2-queue model, we obtain relatively simple explicit expressions for the Laplace transform of the joint buffer content in several special cases. (iii) Asymptotic results are obtained for the 2-queue model when the above-mentioned buffer content removal process approaches a shot-noise process. (iv) For the general <span><math><mi>n</mi></math></span>-queue case, we show how all moments of the buffer contents at all queues can be obtained. (v) For the general <span><math><mi>n</mi></math></span>-queue case, we sketch an approximation method that allows one in principle to derive tractable expressions for the Laplace transform of the buffer content at each queue, with exact mean buffer contents at all queues.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102513"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145417138","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Switching constrained OCO with predictions and feedback delays","authors":"Weici Pan,&nbsp;Zhenhua Liu","doi":"10.1016/j.peva.2025.102524","DOIUrl":"10.1016/j.peva.2025.102524","url":null,"abstract":"<div><div>We examine Online Convex Optimization (OCO) problems with feedback delay and a strict limit on decision switching, which exists in applications such as smart grid and learning. Existing algorithms developed for traditional OCO struggle in this setting, often violating switching constraints or incurring high regrets, as evidenced by simulations. In this paper, we establish a new algorithm, Follow-the-Maximally-Coupled-Latest-Leader (FMCLL), achieving a near-optimal regret of <span><math><mrow><mi>O</mi><mrow><mo>(</mo><mi>T</mi><mo>/</mo><mi>S</mi><mo>)</mo></mrow></mrow></math></span> for such problems with delayed feedbacks and a bound of <span><math><mrow><mi>O</mi><mrow><mo>(</mo><mi>T</mi><mo>/</mo><mi>S</mi><mo>−</mo><mi>τ</mi><mo>)</mo></mrow></mrow></math></span> for problems with predictions of <span><math><mi>τ</mi></math></span> rounds, even though the player is only allowed to move at most <span><math><mi>S</mi></math></span> times in expectation across <span><math><mi>T</mi></math></span> rounds. FMCLL meets performance bounds in scenarios with delays and predictions by using maximal coupling sampling to inform algorithm design for switching-constrained problems. To better apply our framework to practical applications, we also extend the algorithm and results to the bandit feedback setting. Simulations demonstrate FMCLL’s superiority over traditional Gradient Descent or Follow-the-Leader algorithms, excelling under adversarial or stochastic losses and reducing constraint violations.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102524"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145517001","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"LLMEmu: A lightweight performance emulator for high-fidelity distributed LLM training","authors":"Siyuan Yang ,&nbsp;Enda Yu ,&nbsp;Pingjing Lu ,&nbsp;Dezun Dong","doi":"10.1016/j.peva.2025.102526","DOIUrl":"10.1016/j.peva.2025.102526","url":null,"abstract":"<div><div>The prohibitive cost of training trillion-parameter large language models (LLMs) necessitates low-cost emulation tools for distributed system optimization. In modern large-scale clusters, communication often becomes the primary bottleneck to scalability. However, existing emulators, such as vTrain and ASTRA-Sim, overlook dynamic network factors that significantly impact performance at scale, resulting in limited emulation accuracy. This work offers an efficient and reliable tool for training system optimization and parallel strategy exploration, considerably lowering the barrier to large-scale AI research. We present LLMEmu, a distributed training emulator that combines real kernel profiling and actual communication execution. First, computation is profiled through real CUDA kernel traces on GPU nodes to construct an operator-level latency lookup table, enabling GPU-like execution on CPU clusters. Second, inter-node communication is executed using communication library primitives (e.g., AllReduce, Send/Recv), triggered by communication anchors embedded in the execution graph, and implemented using a pluggable communication backend. LLMEmu can seamlessly model hybrid parallelism strategies and supports multiple collective algorithms. Its lightweight design incorporates gradient bucketing with latency reuse to minimize overhead while maintaining extensibility to various network interconnects. The effectiveness of LLMEmu is validated through its performance results, demonstrating an average prediction error of only 2.17% on 24-GPU clusters, which outperforms vTrain by 21.09%, and confirming its scalability in modeling training cost distributions across 128-node CPU emulations under varying network conditions.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102526"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145516998","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Near-optimal PCM wear leveling under adversarial attacks","authors":"Tomer Lange,&nbsp;Joseph (Seffi) Naor,&nbsp;Gala Yadgar","doi":"10.1016/j.peva.2025.102522","DOIUrl":"10.1016/j.peva.2025.102522","url":null,"abstract":"<div><div>Phase-change memory (PCM) is a promising memory technology known for its speed, high density, and durability. However, each PCM cell can endure only a limited number of erase and subsequent write operations before failing, and the failure of a single cell can limit the lifespan of the entire device. This vulnerability makes PCM particularly susceptible to adversarial attacks that induce excessive writes to accelerate device failure. To counter this, wear-leveling techniques aim to distribute write operations evenly across PCM cells.</div><div>In this paper, we study the <em>online PCM utilization problem</em>, which seeks to maximize the number of write requests served before any cell reaches the erase limit. While extensively studied in the systems and architecture communities, this problem remains largely unexplored from a theoretical perspective. We bridge this gap by presenting a novel algorithm that leverages hardware feedback to optimize PCM utilization. We prove that our algorithm achieves near-optimal worst-case guarantees and outperforms state-of-the-art practical solutions both theoretically and empirically, providing an efficient approach to prolonging PCM lifespan.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102522"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145516996","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"γ-CounterBoost: Optimizing response time tails using job type information only","authors":"Nils Charlet,&nbsp;Benny Van Houdt","doi":"10.1016/j.peva.2025.102514","DOIUrl":"10.1016/j.peva.2025.102514","url":null,"abstract":"<div><div>In a recent paper the <span><math><mi>γ</mi></math></span>-Boost scheduling policy was shown to minimize the tail of the response time distribution in a light-tailed M/G/1-queue. This policy schedules jobs using a boosted arrival time, defined as the arrival time of a job minus its boost, where the boost of a job depends on its exact job size. The <span><math><mi>γ</mi></math></span>-Boost policy can also be used when only partial job size information is available, such as the type of an incoming job. In such case the boost <span><math><msub><mrow><mi>b</mi></mrow><mrow><mi>i</mi></mrow></msub></math></span> of a job depends solely on its type <span><math><mi>i</mi></math></span> and <span><math><mi>γ</mi></math></span>-Boost was shown to optimize the tail among all boost policies, where a boost policy is fully determined by the <span><math><msub><mrow><mi>b</mi></mrow><mrow><mi>i</mi></mrow></msub></math></span> values. In the partial information setting <span><math><mi>γ</mi></math></span>-Boost relies on two types of information: job types and arrival times.</div><div>This paper focuses on the problem of minimizing the tail in a light-tailed M/G/1-queue in the partial job size information setting when the scheduler only makes use of the job types and <em>does not exploit arrival times</em>. Prior work showed that in case of 2 job types the so-called Nudge-<span><math><mi>M</mi></math></span> policy minimizes the tail in a large class of scheduling policies. In this paper we introduce the <span><math><mi>γ</mi></math></span>-CounterBoost policy in the partial information setting with <span><math><mrow><mi>d</mi><mo>≥</mo><mn>2</mn></mrow></math></span> job types and prove that it minimizes the tail in an even broader class of scheduling policies called Contextual CounterBoost policies. The <span><math><mi>γ</mi></math></span>-CounterBoost policy reduces to the Nudge-<span><math><mi>M</mi></math></span> policy in case of <span><math><mrow><mi>d</mi><mo>=</mo><mn>2</mn></mrow></math></span> job types.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102514"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145517053","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Response time in a pair of processor sharing queues with Join-the-Shortest-Queue scheduling","authors":"Julianna Bor,&nbsp;Peter G. Harrison","doi":"10.1016/j.peva.2025.102509","DOIUrl":"10.1016/j.peva.2025.102509","url":null,"abstract":"<div><div>Join-the-Shortest-Queue (JSQ) is the scheduling policy of choice for many network providers, cloud servers, and traffic management systems, where individual queues are served under the processor sharing (PS) queueing discipline. A numerical solution for the response time distribution in two parallel PS queues with JSQ scheduling is derived for the first time. Using the generating function method, two partial differential equations (PDEs) are obtained corresponding to conditional response times, where the conditioning is on a particular traced task joining the first or the second queue. These PDEs are functional equations that contain partial generating functions and their partial derivatives, and therefore cannot be solved by commonly used techniques. We are able to solve these PDEs numerically with good accuracy and perform the deconditioning with respect to the queue-length probabilities by evaluating a certain complex integral. Numerical results for the density and the first four moments compare well against regenerative simulation.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102509"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145099156","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"TiFSN: A wavelet-EC-TCN model for quadrotor UAV trajectory prediction based on time–frequency–spatial feature fusion","authors":"Huan Zhao ,&nbsp;Yong Kou ,&nbsp;Yuxin Xue ,&nbsp;Shuang Wang ,&nbsp;Zhaojun Gu","doi":"10.1016/j.peva.2025.102515","DOIUrl":"10.1016/j.peva.2025.102515","url":null,"abstract":"<div><div>Flight trajectory prediction (FTP) with high precision is the core technology for the autonomous flight of quadrotor unmanned aerial vehicles (UAVs) in environments with limited navigation signals. In response to the problem that most existing methods focus on the features of a single domain and ignore the cross-domain feature correlation, making it challenging to maintain high accuracy in FTP, a prediction model based on time–frequency–spatial feature fusion named TiFSN is proposed. Firstly, based on wavelet transform technology, the velocity signal is extended to time–frequency joint features. Furthermore, a fusion mechanism between time–frequency domain features and attitude angles is established, so that a multi-domain feature set with time–frequency–spatial perception can be constructed. Finally, an extended channels-based temporal convolutional network (EC-TCN) is designed, which achieves high-precision FTP by expanding the feature receiving field. Experiments were conducted on real flight datasets, and the results show that the model significantly improved the evaluation metrics compared to baseline methods. The generalization test of various complex FTP tasks using the onboard CPU also verified the excellent performance of the TiFSN. The ablation experiment further revealed the influence of wavelet decomposition depth and the strategy of expanded channels on the performance.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102515"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466726","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"swPredictor: A data-driven performance model for distributed data parallelism training on large-scale HPC clusters","authors":"Xianyu Zhu ,&nbsp;Ruohan Wu ,&nbsp;Junshi Chen ,&nbsp;Hong An","doi":"10.1016/j.peva.2025.102530","DOIUrl":"10.1016/j.peva.2025.102530","url":null,"abstract":"<div><div>Given the complexity of heterogeneous architectures and multi-node collaboration, large-scale HPC (High-Performance Computing) clusters pose challenges in resource utilization and performance optimization during distributed data parallelism (DDP) training. Performance modeling aims to identify application bottlenecks and guide algorithm design, but existing performance models rarely consider the impact of system architecture on communication performance or provide a systematic analysis of distributed training. To address these issues, this paper proposes swPredictor, a data-driven performance model devised for accurately predicting the performance of DDP training. First, an original performance dataset is developed based on various communication patterns at runtime to avoid systematic errors. Subsequently, a novel multi-branch module FNO-Inception is proposed, combining FNO (Fourier Neural Operator) layer with Inception structure to simultaneously utilize various frequency features. Finally, by introducing the FNO-Inception module, a novel regression model FI-Net is constructed to fit complex nonlinear relationships. The experimental results demonstrate that FI-Net can accurately predict the performance of DDP training on the Sunway OceanLight supercomputer with an overall MAPE of 0.93%, which outperforms the other baseline models.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102530"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145516999","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"CommonSense: Efficient Set Intersection (SetX) protocol based on compressed sensing","authors":"Jingfan Meng,&nbsp;Tianji Yang,&nbsp;Jun Xu","doi":"10.1016/j.peva.2025.102520","DOIUrl":"10.1016/j.peva.2025.102520","url":null,"abstract":"<div><div>Set reconciliation (SetR) is an important research problem that has been studied for over two decades. In this problem, two large sets <span><math><mi>A</mi></math></span> and <span><math><mi>B</mi></math></span> of objects (tokens, files, records, etc.) are stored respectively at two different network-connected hosts, which we name Alice and Bob respectively. Alice and Bob need to communicate with each other to learn the set union <span><math><mrow><mi>A</mi><mo>∪</mo><mi>B</mi></mrow></math></span> (which then becomes their reconciled state), at low communication and computation costs. In this work, we study a different problem intricately related to SetR: Alice and Bob collaboratively compute <span><math><mrow><mi>A</mi><mo>∩</mo><mi>B</mi></mrow></math></span>. We call this problem SetX (set intersection). Although SetX is just as important as SetR, it has never been properly studied in its own right. Rather, there is an unspoken perception by the research community that SetR and SetX are equally difficult (in costs), and hence “roughly equivalent.” Our first contribution is to show that SetX is fundamentally a much “cheaper” problem than SetR, debunking this long-standing perception. Our second contribution is to develop a novel SetX solution, the communication cost of which handily beats the information-theoretic lower bound of SetR. This protocol is based on the idea of compressed sensing (CS), which we describe here only for the special case of <span><math><mrow><mi>A</mi><mo>⊆</mo><mi>B</mi></mrow></math></span> (We do have a more sophisticated protocol for the general case). Our protocol is for Alice to encode <span><math><mi>A</mi></math></span> into a CS sketch <span><math><mrow><mi>M</mi><msub><mrow><mi>1</mi></mrow><mrow><mi>A</mi></mrow></msub></mrow></math></span> and send it to Bob, where <span><math><mi>M</mi></math></span> is a CS matrix with <span><math><mi>l</mi></math></span> rows and <span><math><msub><mrow><mi>1</mi></mrow><mrow><mi>A</mi></mrow></msub></math></span> is the binary vector representation of <span><math><mi>A</mi></math></span>. Our key innovation here is to make <span><math><mi>l</mi></math></span> (the sketch size) just large enough (for the sketch) to summarize <span><math><mrow><mi>B</mi><mo>∖</mo><mi>A</mi></mrow></math></span> (what Alice misses). In contrast, in existing protocols <span><math><mi>l</mi></math></span> needs to be large enough to summarize <span><math><mi>A</mi></math></span> (what Alice knows), which is typically much larger in cardinality. Our third contribution is to design a CS matrix <span><math><mi>M</mi></math></span> that is both “friendly” to (the performance of) applications and “compliant” with CS theory.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102520"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145517051","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Mitigating massive access with Quasi-Deterministic Transmission: Experiments and stationary analysis","authors":"Jacob Bergquist ,&nbsp;Erol Gelenbe ,&nbsp;Mohammed Nasereddin ,&nbsp;Karl Sigman","doi":"10.1016/j.peva.2025.102512","DOIUrl":"10.1016/j.peva.2025.102512","url":null,"abstract":"<div><div>The Massive Access Problem arises due to devices that forward packets simultaneously to servers in rapid succession, or by malevolent software in devices that flood network nodes with high-intensity traffic. To protect servers from such events, attack detection (AD) software is installed on servers, and the Quasi-Deterministic Transmission Policy (QDTP) has been proposed to “shape traffic” and protect servers, allowing attack detection to proceed in a timely fashion by delaying some of the incoming packets individually based on their arrival times. QDTP does not cause packet loss, and can be designed so that it does not increase end-to-end packet delay. Starting with measurements taken on an experimental test-bed where the QDPT algorithm is installed on a dedicated processor, which precedes the server itself, we show that QDPT protects the server from attacks by accumulating arriving packets at the input of the QDTP processor, then forwarding them at regular intervals to the server. We compare the behaviour of the server, with and without the use of QDTP, showing the improvement it achieves, provided that its “delay” parameter is correctly selected. We analyze the sample paths associated with QDTP and prove that when its delay parameter is chosen in a specific manner, the end-to-end delay of each packet remains unchanged as compared to an ordinary First-In-First-Out system. An approach based on stationary ergodic processes is developed for the stability conditions. Assuming mutually independent and identically distributed inter-arrival times, service times and QDTP delays, we exhibit the positive recurrent structure of a two-dimensional Markov process and its regeneration points.</div></div>","PeriodicalId":19964,"journal":{"name":"Performance Evaluation","volume":"170 ","pages":"Article 102512"},"PeriodicalIF":0.8,"publicationDate":"2025-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145466621","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}