high-performance and large-scale systems

Selecting appropriate parallel programming models for heterogeneous HPC architectures remains challenging due to divergent hardware characteristics and software trade-offs. Method: This paper conducts the first multi-dimensional quantitative comparison of MPI, OpenMP, and CUDA—evaluating architectural adaptability, scalability bottlenecks, development complexity, and domain suitability—and proposes a hybrid programming model selection framework tailored to heterogeneity. The framework integrates communication modeling, memory contention analysis, and GPU kernel optimization for empirical validation. Contribution/Results: Experiments show MPI achieves >92% strong scaling efficiency in distributed, communication-intensive workloads; OpenMP delivers 3.8× speedup on shared-memory loop-parallel tasks; CUDA attains up to 12.5× acceleration on data-parallel kernels; and hybrid strategies yield an average 27% improvement in end-to-end performance. The study provides both theoretical foundations and practical guidelines for optimizing and co-designing programming models in heterogeneous HPC environments.

OpenMP’s static scheduling policies rely heavily on manual tuning and struggle to adapt to diverse applications and heterogeneous platforms. To address this, we propose a dynamic scheduling selection framework that synergistically integrates domain expertise with reinforcement learning (RL). At runtime, the framework adaptively selects optimal scheduling policies based on real-time application characteristics and system state, supports cross-platform deployment, and generalizes to multi-level parallelism (e.g., MPI+OpenMP). Evaluated across six representative applications and three hardware platforms, our hybrid approach outperforms conventional static schedulers—achieving an average 23.6% speedup and reducing load imbalance by 41.2%. The key innovation lies in embedding expert knowledge into RL’s reward formulation and action-space constraints, thereby enabling interpretable, efficient, and generalizable dynamic scheduling optimization.

Traditional operating systems suffer from poor scalability on many-core processors and low parallel efficiency due to their inability to perceive application semantics. To address this, we propose NetworkedOS—a novel application-aware, networked OS architecture. Our approach leverages compile-time dynamic instruction dependency analysis to construct a multi-layer network model that explicitly captures runtime dependencies among applications, the kernel, and hardware. We further design an overlapping graph partitioning algorithm to jointly optimize parallel execution and inter-core communication overhead, and implement a runtime process affinity mapping scheduler. Crucially, NetworkedOS breaks the conventional “black-box” OS assumption regarding application semantics for the first time. Experimental evaluation shows that NetworkedOS achieves a 7.11× speedup over Linux on a 128-core system and a 2.01× improvement over Barrelfish on a 64-core system, significantly enhancing scalability and resource utilization under large-scale parallel workloads.

This work addresses the growing diversity of GPU workloads in high-performance computing (HPC) clusters, where traditional per-application profiling incurs high overhead and poor scalability, hindering joint optimization of performance and power. To overcome these limitations, the authors propose Minos—the first unified GPU workload classification framework based on low-overhead feature extraction. By integrating lightweight runtime profiling, feature engineering, and clustering analysis, Minos groups behaviorally similar workloads into a limited set of categories, drastically reducing profiling costs for new applications. Experimental evaluation across 18 representative graph analytics, HPC, and machine learning workloads demonstrates that Minos achieves average prediction errors of only 4% for power and 3% for performance—outperforming the state-of-the-art by 10%. Furthermore, it reduces profiling time for frequency-limiting behavior of unseen applications by 89%, substantially improving cross-domain generalization.

This paper addresses the lack of clarity regarding the diversity and evolutionary trajectories of modern workload schedulers. We propose a cross-layer taxonomy comprising three categories: OS process scheduling, cluster job scheduling, and big-data scheduling. Through algorithmic feature analysis and historical comparative study, we systematically characterize the design rationales, optimization objectives, and technological evolution of these schedulers, uncovering shared design patterns across local and distributed environments. Our key contribution is the first unified classification framework, which identifies three fundamental differentiating dimensions: resource abstraction granularity, scheduling timing, and feedback mechanism. Based on this analysis, we distill general-purpose scheduling design principles targeting heterogeneity, scalability, and QoS guarantees. The study provides both theoretical foundations and practical guidance for scheduler selection, cross-layer coordination optimization, and next-generation scheduler architecture design.

This work addresses the high complexity of existing high-performance computing (HPC) performance analysis tools, which hinders students’ intuitive understanding of parallel program performance issues. To bridge this gap, the paper introduces EduMPI—the first educational tool that integrates HPC cluster operations and MPI performance analysis within a streamlined graphical interface. EduMPI enables near real-time, physically node-layout-aware communication visualization, facilitating interactive identification of load imbalance and other performance bottlenecks. User studies demonstrate that, compared to professional-grade tools, EduMPI significantly lowers the learning barrier and effectively enhances students’ comprehension of parallel performance characteristics, thereby improving the practicality and accessibility of parallel programming education.

To address the challenge in cloud environments where HPC applications struggle to simultaneously achieve high resource utilization and low response latency for high-priority jobs—given that traditional MPI lacks automatic elasticity and existing elastic frameworks (e.g., Charm++) require manual application refactoring—this paper proposes the first Kubernetes Operator-based elastic HPC job scheduler. It innovatively integrates Charm++’s migratable object model with Kubernetes’ native orchestration capabilities, enabling fully automated, application-transparent dynamic scaling without source-code modifications. The scheduler employs a priority-aware elasticity algorithm to adjust resource allocation in real time within containerized environments. Experimental results demonstrate negligible scaling overhead, a 32% increase in average cluster resource utilization, and a 47% reduction in average response time for high-priority jobs—significantly outperforming static scheduling baselines.

This paper addresses the lack of systematic optimization for CPU and memory resource allocation during the Release phase of cloud-native DevOps. We propose the first pre-deployment offline performance optimization framework for microservices—distinct from mainstream auto-scaling research focused on the Ops phase. Our approach performs fine-grained resource configuration tuning *before* deployment, thereby mitigating auto-scaling failures caused by suboptimal memory provisioning. Methodologically, we integrate Bayesian optimization, statistical experimental design, and a goal-directed factor screening strategy to balance sampling cost and approximation accuracy. Extensive evaluation on the TeaStore benchmark demonstrates that our pre-deployment optimization significantly improves memory suitability and API-level resource utilization. Moreover, it empirically validates the necessity and context-dependent applicability of factor screening under varying optimization objectives.

Scheduling data-intensive workloads in large-scale distributed systems faces challenges including complexity, heterogeneous parallelism, data locality constraints, and multi-dimensional QoS optimization (e.g., timeliness, fault tolerance, energy efficiency). Method: This paper proposes a novel workload classification scheme grounded in data characteristics and service requirements; systematically surveys and structures mainstream scheduling strategies, exposing critical limitations in dynamic adaptability, fine-grained fault tolerance, and energy–QoS co-optimization; and introduces a unified scheduling framework integrating data-locality awareness, elastic parallel scheduling, QoS-tiered guarantees, and energy-aware resource allocation. Contribution/Results: The study establishes a scalable classification paradigm, delivers a clear technology evolution roadmap, and identifies a prioritized list of open research challenges—thereby advancing foundational understanding and guiding future design of intelligent, holistic schedulers for modern distributed data systems.

This study addresses the lack of systematic comparison between asynchronous multitasking (AMT) runtime systems and MPI in terms of both performance and programming productivity. Leveraging the Task Bench framework, it presents the first unified benchmark incorporating Itoyori and ItoyoriFBC alongside HPX and MPI, evaluating them across diverse workloads using PGAS abstractions, RDMA-based work stealing, and future-based synchronization. Quantitative analysis via application efficiency, METG, lines of code, and library constructs reveals distinct trade-offs: Itoyori achieves the highest efficiency with the most concise code; MPI excels in regular, low-communication tasks yet requires verbose implementations; HPX demonstrates robust stability but lowest productivity; and ItoyoriFBC offers enhanced expressiveness at a modest performance cost.