Structured mesh generation serves as a crucial preprocessing step in numerical simulations and can be formulated as a mapping problem from geometry to structured mesh. Existing approaches typically establish an isolated mapping for each geometry. This geometry-specific paradigm fails to capture and leverage commonalities across geometries, inevitably requiring recomputation or costly retraining for new geometries. To overcome this limitation, we propose ICL-Mesh, a meta-learning framework based on in-context learning (ICL) for structured mesh generation. It treats learning one mapping as one task and trains a single neural network to extract commonalities across tasks and learn from in-context examples within each task, enabling rapid generalization to unseen tasks without parameter updates. Experimental results demonstrate that ICL-Mesh effectively generalizes to diverse geometries with only a few context examples, and even without examples. It also exhibits robustness to in-context example order sensitivity and can be extended to various mesh generation scenarios, including mesh refinement and coarsening.

While approximate computing offers a promising paradigm for performance gains, significant technical challenges obstruct its practical adoption in High Performance Computing (HPC). The process of manually identifying approximable regions, quantifying the risk of cascading errors, and navigating the vast combinatorial tuning space is intricate and error-prone. To overcome these obstacles, we present Tianhe Approximate Computing (THAC), a framework that establishes a systematic, automated methodology. THAC implements a principled, three-stage workflow: (1) a hybrid AI-driven synthesizer that automatically identifies potential approximation candidates; (2) a principled Uncertainty Quantification screen that rigorously validates and prunes high-risk options through a principled risk-aware screening stage; and (3) a novel hierarchical Bayesian optimizer that efficiently navigates the search space. Although demonstrated on OpenMP as a representative case study, the framework is designed with extensibility for generic parallel patterns. Evaluation on a diverse suite of scientific benchmarks (including computational fluid dynamics and hydrodynamics) demonstrates the effectiveness of THAC, achieving a geometric mean speedup of 2.1× under strict quality budgets. These results confirm THAC as a robust solution that makes approximate computing in HPC practical and reliable.

Krylov subspace methods are widely used in scientific computing to solve large sparse linear systems and eigenvalue problems. Their performance bottleneck is often dominated by high-order matrix-power kernels (MPK), especially in polynomial preconditioners that must scale to millions or billions of variables. We present Diagonal Block MPK (DBMPK), a lightweight and parallel-friendly optimization that partitions the input matrix into diagonal blocks and off-diagonal regions. This design enables efficient intra-block data reuse and eliminates inter-block dependencies. It improves cache locality, parallelism, and reduces preprocessing overheads, compared to existing techniques. Our evaluation on x86 and Arm HPC platforms shows that DBMPK improves MPK performance by 26.6%-38.4%. When applied to polynomial preconditioners for linear systems and eigenvalue problems, it achieves consistent end-to-end speedups of 18.6%-34.0%, including in weak scaling tests on 128 nodes, demonstrating strong scalability and practical impact.

Sparse matrix–matrix multiplication (SpMM) is a fundamental operation in scientific computing with broad applications across numerous domains. Tiling is a key optimization technique for improving data locality and is widely adopted in high-performance computing. However, the irregular data access patterns inherent to SpMM make it challenging to exploit tiling effectively for data reuse. In this article, we propose MaSpMM , a memory-aware SpMM framework that integrates cache-aware tiling with a segment-oriented data layout. MaSpMM stores matrices as continuous segments to enhance data locality within each tile. Moreover, since many sparse matrices in real-world applications exhibit symmetry, we further develop MaSpMM-Sym, an extension that recursively partitions symmetric matrices to eliminate write conflicts and further improve locality. To adapt to diverse scenarios, we finally introduce MaSpMM-Adap, which adaptively selects the most suitable approach for each input matrix. Comprehensive evaluations on both x86 and ARM CPUs demonstrate that MaSpMM-Adap achieves average speedups of up to 1.86× over Intel oneMKL, 1.84× over ASpT, and 1.75× over J-Stream.

Industrial image anomaly detection (IAD) is a pivotal topic with huge value. Due to the nature of anomalies, real anomalies in a specific modern industrial domain (i.e., domain-specific anomalies) are usually too rare to collect, which severely hinders IAD. Thus, zero-shot anomaly synthesis (ZSAS), which synthesizes pseudo anomaly images without any domain-specific anomaly, emerges as a vital technique for IAD. However, existing solutions are either unable to synthesize authentic pseudo anomalies, or require cumbersome training. Thus, we focus on ZSAS and propose a brand-new paradigm that can realize both authentic and training-free ZSAS. It is based on a chronically-ignored fact: Although domain-specific anomalies are rare, real anomalies from other domains (i.e., cross-domain anomalies) are actually abundant and directly applicable to ZSAS. Specifically, our new ZSAS paradigm makes three-fold contributions: First, we propose a novel method named Cross-domain Anomaly Injection (CAI), which directly exploits cross-domain anomalies to enable highly authentic ZSAS in a training-free manner. Second, to supply CAI with sufficient cross-domain anomalies, we build the first Domain-agnostic Anomaly Dataset (DAAD) within our best knowledge, which provides ZSAS with abundant real anomaly patterns. Third, we propose a CAI-guided Diffusion Mechanism, which can further break the quantity limit of real anomalies and enable unlimited anomaly synthesis. Our head-to-head comparison with existing ZSAS solutions justifies the superior performance of our paradigm for IAD and demonstrates it as an effective and pragmatic ZSAS solution.

Text-to-image person re-identification (TIReID) aims to retrieve semantically related images from a large gallery given a text query. Most existing TIReID methods train the model on an ideal assumption that positive (negative) samples are semantically correlated (uncorrelated) on the visual and textual modal. However, we observe that incorrect annotation is unavoidable and ambiguous textual description is often taken as the input in practice, which leads to false negatives that ruin the feature aligning between modals for deviated identification. This work presents a novel False Negative Mitigation (FNM) method by identifying potential false negatives through distribution differences and mitigating them with dedicated loss. Specifically, the false negative mitigation loss is designed to adaptively adjust the optimization margin for potential false negatives in the latent space. Moreover, a Dual-Level Feature Representation method is introduced to leverage both global and local features during the mitigation, whereas a Momentum Contractive (MoC) module is plugged to enrich the training data for accurate similarity distribution estimation. We conduct extensive experiments on three public benchmark datasets and the results demonstrate that our FNM method achieves State-of-The-Art (SoTA) performance at all evaluation metrics.

This paper focuses on optimizing the Matrix-Power Kernel (MPK), which relies on a series of Sparse Matrix-Vector multiplications (SpMVs) using the same sparse matrix. MPK is a crucial component of Krylov subspace methods for solving large sparse linear systems in various fields, including circuit simulations. MPK offers a potential for matrix reuse in cache, which can accelerate memory-bound sparse solvers. Additionally, many sparse matrices encountered in applications are symmetric, allowing us to reduce the memory footprint for SpMVs by half. However, reusing the matrix introduces data dependencies between subsequent SpMVs, and symmetric SpMVs can result in data conflicts during shared-memory parallelization. Previous research has often focused on either matrix reuse or symmetry, failing to leverage both aspects effectively. This paper proposes a unified, memory-efficient approach called Me-MPK that takes advantage of both cache reuse and matrix symmetry for MPK on shared-memory multi-core systems. We first introduce a unified dependency graph for a sparse matrix, which represents all potential data dependencies and conflicts. Next, we perform architecture-aware recursive partitioning on this graph to create subgraphs and formulate a separating subgraph that decouples all dependencies and conflicts among the subgraphs. These independent subgraphs are then scheduled for parallel execution of SpMV or symmetric SpMV in a specified order to optimize cache reuse. We apply Me-MPK in two s-Step Krylov subspace solvers, and our evaluations show that Me-MPK significantly outperforms the current state-of-the-art solutions, delivering an average speedup of up to 2.00X and 1.86X on X86 and ARM CPUs, respectively. As a result, we achieve overall speedup in the sparse solvers of up to $\mathbf{1. 6 5 X}$ and $\mathbf{1. 5 8 X}$.

With the rapid growth of multimodal and multi-view data, multi-view bipartite graph clustering (MVBGC) has emerged as a promising solution for large-scale tasks, which with linear complexity. However, most methods adhere to a unidirectional “encoding” design, where the bipartite graph is directly constructed from input data. Enlightened by the prevalent encoding-decoding in deep learning, this paper rethinks existing paradigms and proposes a novel “auto-encoding” MVBGC framework, named BGAE. Our model seamlessly integrates encoding, bipartite graph learning, and decoding modules within a self-supervised learning framework. The encoding module extracts a joint representation from input data, the bipartite graph learning module learns a discriminative bipartite graph in latent semantic space, and the decoding module reconstructs the input data by the structural information. Extensive experiments verify the superiority of our novel design, particularly highlighting the critical role of “decoding” learning. This work represents the first attempt to explore encoding-decoding design in MVBGC.

3D medical image classification is challenging due to small, subtle lesions and substantial irrelevant context, which often mislead deep models. Inspired by the top-down diagnos-tic process of clinicians—first identifying anatomical context, then locating anomalies—we propose Prompt-Guided 3D Vision Transformer (PG 3D- ViT), a framework that simulates clinical reasoning through prompt-driven attention. To address limited 3D training data, PG3D- ViT leverages 2D masked auto encoder (MAE) pretraining to learn transferable image features. Through the prompt generation module, consistency difference analysis is performed between normal and abnormal samples to extract anatomical structure and global spatial prompt information related to the lesion context. These prompts are injected as query into a cross-attention mechanism, guiding the model to focus on lesion-relevant regions across the 3D volume. Evaluated on 7 public datasets spanning multiple modalities and pathologies, PG3D-ViT achieves a 1.88% average AUC improvement over state-of-the-art methods. The attention map visualizations demonstrate that the model can accurately localize lesion regions, validating the effectiveness of the clinical prompting mechanism in enhancing both the performance and interpretability of 3D medical image classification. The code is available at the provided link11https://github.comJUMED-P/PG3D-ViT

Text-Image Person Re-identification (TIReID) aims to retrieve the relevant pedestrian images according to the given textual query. Recent methods typically achieve this goal through image-text contrastive learning, which assumes that only paired images and texts from the same pedestrian are considered positive samples. However, we observe that there exist negative samples, termed false negatives, that are highly semantically related to the anchor in practice. Training with these false negatives may adversely affect feature representation learning and semantic alignment between modalities. This work proposed a false negative detection and suppression (FNDS) method to mitigate their adverse impact. Our FNCD consists of a False Negative Consensus Detection (FNCD) mechanism and an Adaptive False Negative Suppression (AFNS) method. FNCD combines dual-grained detection to consensually identify potential false negatives, while AFNS assigns adaptive weights to the false negative similarities for more robust suppression. Extensive experiments conducted on three public benchmark datasets demonstrate the effectiveness of the proposed method.

Implicit neural representation (INR) has now been thrust into the limelight with its flexibility in high-fidelity flow field reconstruction tasks. However, the lack of standard benchmarking datasets and the grid independence assumption for INR-based methods hinder progress and adoption in real-world simulation scenarios. Moreover, naive adoptions of existing INR frameworks suffer from limited accuracy in capturing fine-scale structures and spatiotemporal dynamics. Tacking these issues, we first introduce HFR-Beach, a 5.4 TB public large-scale CFD dataset with 33,600 unsteady 2D and 3D vector fields for reconstructing high-fidelity flow fields. We further present PEINR, a physics-enhanced INR framework, to enrich the flow fields by concurrently enhancing numerical-precision and grid-resolution. Specifically, PEINR is mainly composed of physical encoding and transformer-based spatiotemporal fuser (TransSTF). Physical encoding decouples temporal and spatial components, employing Gaussian coordinate encoding and localized encoding techniques to capture the nonlinear characteristics of spatiotemporal dynamics and the stencil discretization of spatial dimensions, respectively. TransSTF fuses both spatial and temporal information via transformer for capturing long-range temporal dependencies. Qualitative and quantitative experiments and demonstrate that PEINR outperforms state-of-the-art INR-based methods in reconstruction quality.

The Sparse General Matrix-Matrix Multiplication (SpGEMM) is widely used in a variety of applications. However, research on optimizing SpGEMM for high-performance digital signal processors (DSPs) has been limited. We present VES, a method to accelerate SpGEMM on multi-core DSPs, using the FT-M7032 platform as a case study. Based on the ESC algorithm, VES enhances computational efficiency through vectorized expansion operations, a double-buffering strategy, and an optimized vectorized sorting method. We provide an in-depth analysis of the bottlenecks in vectorized sorting and introduce an efficient vectorized reduction method that significantly improves instruction-level pipeline throughput. Experimental results show that VES outperforms existing methods HASH, ESC, and SPA by an average of 1.12×, 1.90×, and 22.34× on 1,931 sparse matrices, with maximum speedups of 8.69×, 65.9×, and 821.6×, respectively.

Incomplete Cholesky (IC) factorization is widely used to precondition large sparse positive definite symmetric linear equations, since it can effectively reduce the number of iterations and enhance the solution efficiency compared to the unconditioned Conjugate Gradient (CG) iterative algorithm. In this paper, we have carried out optimization research on the IC factorization parallel algorithms based on the MIMD many-core architecture PEZY-SC3s processor. We propose a parallel algorithm for IC factorization under the MIMD architecture. This algorithm is based on the level-scheduling and graph coloring reordering algorithm. We optimize algorithm performance using various measures, including memory access optimization based on vector units and on-chip local memory, load balancing optimization based on dynamic thread scheduling, and task allocation optimization based on graph partitioning. Experimental results demonstrate that our IC parallel factorization achieves average speedups of 17.5x and 8.4x over the two cuSPARSE implementations on an NVIDIA A30 GPU, and 230.6x and 83.0x over Ginkgo’s serial and OpenMP implementations on a dual-socket Intel Xeon 4314 platform. Our study demonstrates that the PEZY architecture with effective algorithmic design still achieves good performance gains while maintaining low power consumption.

Large-scale graph tasks often have to be conducted in parallel on partitioned graphs.However, current partitioning methods, designed for smaller-scale clusters with a limited number of computing nodes, struggle to scale effectively due to their inability to handle messages transferred through traditional partitioning grids.We present YH-Light, an hierarchy-aware partitioning engine on Tianhe supercomputers to minimize communication.The key idea of YH-Light is to take advantage of the hierarchical communication topology to perform a graph partition based on communication hierarchies, where scattered messages are (i) clustered with hub vertices according to the organization of the computing nodes and (ii) grouped and then exchanged messages according to hierarchical communication domains.We demonstrate YH-Light's effectiveness with synthetic benchmarks and real-world graphs.In particular, the YH-Light-based Graph 500 tests on the Tianhe supercomputer outperform the leading systems in the latest Graph 500 list.Furthermore, YH-Light significantly advances graph processing, surpassing the current state-of-the-art graph partitioning engines and graph systems by orders of magnitude.

Algebraic multigrid (AMG) is widely used to accelerate largescale sparse linear solvers.In distributed environments, neighboring communication overhead in AMG significantly impacts overall solution time.We propose Communication-Reduced Algebraic Multigrid (CRAMG) methods to minimize inter-process data exchange and message count by fusing interpolation/restriction operators with residual computations.This reduces communication frequency from four per level to as few as two.Experiments show up to 45% reduction in data exchange and 35% fewer messages.Performance evaluations on an Intel platform demonstrate significant improvements

Accurate spatio-temporal prediction of oil-water two-phase flow under heterogeneous reservoir conditions is crucial for optimizing oil recovery strategies. Pioneering studies have begun to explore intelligent methods to reduce the computational cost in traditional numerical simulation. However, existing intelligent methods suffer from error accumulation problems, leading to inaccurate evolution results over time. In this paper, we propose CASTNet, a novel spatio-temporal model for predicting the evolution of oil-water two-phase flow. The model designs an encoder-decoder structure to capture and simulate the spatio-temporal variations of oil-water two-phase flow. We design an attention-enhanced ConvLSTM module to strengthen the identification of heterogeneous regions, tackling the issue of failing to detect spatial information in heterogeneous reservoirs. Experiments demonstrate that CASTNet achieves accurate and efficient spatio-temporal prediction of oil-water two-phase flow. It outperforms existing intelligent methods by maintaining a maximum absolute error of 0.08 during long-term simulation, while it achieves one order-of-magnitude improvement in computational efficiency compared to traditional numerical methods.

Deploying convolutional neural networks on low-resource mobile devices for single image super-resolution (SISR) faces the issue of how to balance the parameter amount and performance. The default solution is simultaneously condensing both hierarchical representation and attention features into their respective light proxies. The insight underlying this solution lies in the fact that features are redundant since the super-resolution needs plenty of similar pixels. This work takes it to the next step from the viewpoint of informativeness and discrimination. In detail, we propose an informative disrcimination network (IDNet) for SISR. For informativeness, a multi-scale residual block (MRB) is explored to capture informative spatial details via the scale-in-scale structure. It mines rich intra-layer spatial details based on inter-layer ones of the default hierarchical representation. However, it also incurs feature redundancy. Though attention serves to reduce this redundancy, feature discrimination and pixel-wise structural preservation cannot be guaranteed. Here spatial discrimination attention behaves like the biased discriminant classifier to induce spatial discrimination, while the nuclear-norm regularization recovers the image low-rank structure to reduce artifacts or noises. Importantly, no extra network weights are introduced for model efficiency. Experiments show that IDNet delivers sound performance with fewer parameters, as compared to its cousins.

Solving partial differential equations is a key focus of research in scientific computing. Traditional neural operator methods often face challenges in capturing both global features and local features simultaneously, limiting their effectiveness in complex scenarios. In this paper, we propose the Dual-Spectral Neural Operator (DSNO), an innovative method that efficiently addresses these challenges by integrating both Fourier and wavelet transforms. DSNO employs a dynamic adjustment mechanism, incorporating a hybrid multiplicative-additive fusion strategy to optimally balance the contributions of Fourier and wavelet transforms, thereby enhancing its capacity to capture both global and local features. Extensive experiments demonstrate that our method outperforms existing neural operator methods, achieving state-of-the-art predictive performance on a range of challenging tasks.

Support Vector Machine (SVM) is highly effective at processing high-dimensional, nonlinear data. However, more than 90 % of the training time is spent on kernel matrix calculations, and existing approaches encounter challenges in adapting to heterogeneous architectures. This paper presents an adaptive dual-threshold method leveraging heterogeneous collaboration to accelerate kernel matrix computations. We also analyze the impact of the working set size on training time and accuracy in ThunderSVM to minimize training time. Tasks with distinct characteristics are allocated to appropriate computing cores through heterogeneous collaboration, with dynamic load balancing via adaptive dual thresholds. On a CPU-DSP heterogeneous platform, our method delivers an average speedup of$5.52 \times$compared to the optimal CPU-only implementation.

Partial differential equations (PDEs) form the mathematical foundation for modeling physical systems in science and engineering, where numerical solutions demand rigorous accuracy-efficiency tradeoffs. Mesh movement techniques address this challenge by dynamically relocating mesh nodes to rapidly-varying regions, enhancing both simulation accuracy and computational efficiency. However, traditional approaches suffer from high computational complexity and geometric inflexibility, limiting their applicability, and existing supervised learning-based approaches face challenges in zero-shot generalization across diverse PDEs and mesh topologies. In this paper, we present an $\textbf{U}$nsupervised and $\textbf{G}$eneralizable $\textbf{M}$esh $\textbf{M}$ovement $\textbf{N}$etwork (UGM2N). We first introduce unsupervised mesh adaptation through localized geometric feature learning, eliminating the dependency on pre-adapted meshes. We then develop a physics-constrained loss function, M-Uniform loss, that enforces mesh equidistribution at the nodal level. Experimental results demonstrate that the proposed network exhibits equation-agnostic generalization and geometric independence in efficient mesh adaptation. It demonstrates consistent superiority over existing methods, including robust performance across diverse PDEs and mesh geometries, scalability to multi-scale resolutions and guaranteed error reduction without mesh tangling.

Graph partitioning is essential for effectively managing multi-trillion-edge graphs in distributed computing systems, particularly those spanning hierarchical architectures with thousands of computing nodes. Traditional partitioning strategies neglect hierarchical communication variances across modern high-performance computing (HPC) systems, leading to prohibitive cross overhead when processing trillion-scale graphs. We propose TianheEngine, a hierarchy-aware adaptive partitioning system that takes advantage of the communication hierarchy of the underlying distributed computing system and the sparsity characteristics of the input graphs to improve communication efficiency. We evaluated TianheEngine on fundamental graph operations using both synthetic and real-world datasets. Our extensive experiments use up to 79,024 computing nodes and over 1.2 million processor cores. Experimental results show that TianheEngine is superior to state-of-the-art graph partitioning methods and parallel graph systems and outperforms top-ranked systems on the latest Graph 500 list.

This paper presents DAS-ILU, a Distributed Asynchronous parallel Incomplete LU factorization method based on domain decomposition. DAS-ILU partitions the computational domain into independently processed interior nodes and asynchronously updated separator nodes, thereby reducing cross-processor dependencies and halving the separator size compared to conventional methods. To further improve performance, it employs optimized data exchange patterns to minimize communication overhead and extends support to block-structured sparse matrices via exact block inversions. Comprehensive evaluations on a range of problem types—including structural mechanics, computational fluid dynamics, and reservoir simulation demonstrate the superior performance of DAS-ILU. Compared to state-of-the-art ILU implementations, DAS-ILU achieves solve time speedups of up to 2.07 × over Chow-Patel’s fine-grained parallel ILU and up to 4.11 × over HYPRE’s ILU. Moreover, DAS-ILU exhibits strong robustness when applied to challenging nonsymmetric and indefinite systems.

Efficient processing of large-scale graphs with billions to trillions of edges is essential for training graph-based large language models (LLMs) in web-scale systems. The increasing complexity and size of these models create significant communication challenges due to the extensive message exchanges required across distributed nodes. Current graph engines struggle to effectively scale across hundreds of computing nodes because they often overlook variations in communication costs within the interconnection hierarchy. This paper presents GraphCom, a communication-efficient message graph engine for graph processing on supercomputers. Our key idea is to leverage the network topology information to perform communication hierarchy-aware message aggregation, where messages are (i) gathered to the responsible nodes (referred to as monitors) in the source domains, (ii) transferred between monitors, and (iii) scattered to the target nodes in the target domains. GraphCom's aggregation is more aggressive in that each source domain (instead of the source node). We have implemented GraphCom on top of MPI. We demonstrate GraphCom's effectiveness with synthetic benchmarks and real-world graphs, utilizing up to 79,024 nodes and over 1.2 million processor cores, demonstrating that GraphCom surpasses leading graph- parallel systems and state-of-the-art counterparts in both throughput and scalability. Moreover, we have deployed GraphCom on a production supercomputer, where it consistently outperforms the top solutions on the Graph500 list. These results highlight the potential GraphCom has to significantly improve the efficiency of distributed large-scale graph-based LLM training by optimizing communication between distributed systems, making it an invaluable graph engine for distributed training tasks on web-scale graphs.

Graph data processing is essential for web-scale applications, including social networks, recommendation systems, and web of things (WoT) systems, where large, sparsely connected graphs dominate. Traditional sparse matrix storage formats like compressed sparse row (CSR) face significant memory and performance bottlenecks in distributed, federated, and edge-based computing environments, which are increasingly central to the web. To address this challenge, we propose GraphCSR, a novel storage format that clusters vertices with identical edge degrees and stores only the starting index of each group. This approach minimizes memory overhead and facilitates batch memory access while enhancing overall performance, making it particularly suitable for federated systems and resource-constrained edge nodes. Our experiments across various graph operations and large datasets show that GraphCSR achieves considerable memory savings and performance gains of large-scale, distributed graph processing. When deployed GraphCSR on two production-grade supercomputers, demonstrating its potential for scaling web and WoT graph processing in large-scale distributed computing systems.

Mesh generation is a crucial step in numerical simulations, significantly impacting simulation accuracy and efficiency. However, generating meshes remains time-consuming and requires expensive computational resources. In this paper, we propose a novel method, 3DMeshNet, for three-dimensional structured mesh generation. The method embeds the meshing-related differential equations into the loss function of neural networks, formulating the meshing task as an unsupervised optimization problem. It takes geometric points as input to learn the potential mapping between parametric and computational spaces. After suitable offline training, 3DMeshNet can efficiently output a three-dimensional structured mesh with a user-defined number of quadrilateral/hexahedral cells through the feed-forward neural prediction. To enhance training stability and accelerate convergence, we integrate loss function reweighting through weight adjustments and gradient projection alongside applying finite difference methods to streamline derivative computations in the loss. Experiments on different cases show that 3DMeshNet is robust and fast. It outperforms neural network-based methods and yields superior meshes compared to traditional mesh partitioning methods. 3DMeshNet significantly reduces training times by up to 85% compared to other neural network-based approaches and lowers meshing overhead by 4 to 8 times relative to traditional meshing methods.

Missing traffic data caused by sensor failures or communication errors significantly hinders the efficiency of downstream tasks in Intelligent Transportation Systems (ITS), such as the critical functions of traffic monitoring and decision-making. Since missing data contains important information, it is essential to extract dynamic spatial-temporal correlations in traffic processes by incorporating these missing features. Motivated by these concerns, a novel Dynamic Spatial-Temporal Imputation Network with Missing Features (DSTMIN) is proposed to accurately impute traffic data. DSTMIN comprises an embedding layer, a Mask Attention module (MA), and a Fusion Graph Convolution module (FGC). Specifically, an embedding layer is designed to accurately represent the distribution of missing data, thereby capturing both temporal features and missing features. Furthermore, in order to effectively capture the temporal correlations, MA integrates the missing features to emphasize the significance of observed data and reduce the adverse effects caused using incomplete data. To capture spatial correlations, FGC constructs the spatial graphs and model dynamic spatial correlations from traffic subsequences and the missing-data graph in the presence of missing features. The proposed DSTMIN is adequately evaluated to demonstrate its superior performance on two datasets, which achieves a remarkable 20% reduction in imputation error compared to state-of-the-art methods.

Turbulence, a complex multi-scale phenomenon inherent in fluid flow systems, presents critical challenges and opportunities for understanding physical mechanisms across scientific and engineering domains. Although high-resolution (HR) turbulence data remain indispensable for advancing both theoretical insights and engineering solutions, their acquisition is severely limited by prohibitively high computational costs. While deep learning architectures show transformative potential in reconstructing high-fidelity flow representations from sparse measurements, current methodologies suffer from two inherent constraints: strict reliance on perfectly paired training data and inability to perform multi-scale reconstruction within a unified framework. To address these challenges, we propose HADF, a hash-adaptive dynamic fusion implicit network for turbulence reconstruction. Specifically, we develop a low-resolution (LR) consistency loss that facilitates effective model training under conditions of missing paired data, eliminating the conventional requirement for fully matched LR and HR datasets. We further employ hash-adaptive spatial encoding and dynamic feature fusion to extract turbulence features, mapping them with implicit neural representations for reconstruction at arbitrary resolutions. Experimental results demonstrate that HADF achieves superior performance in global reconstruction accuracy and local physical properties compared to state-of-the-art models. It precisely recovers fine turbulence details for partially unpaired data conditions and diverse resolutions by training only once while maintaining robustness against noise.

In computational fluid dynamics (CFD), mesh-smoothing methods are widely used to refine the mesh quality for achieving high-precision numerical simulations. Specifically, optimization-based smoothing is used for high-quality mesh smoothing, but it incurs significant computational overhead. Pioneer works have improved its smoothing efficiency by adopting supervised learning to learn smoothing methods from high-quality meshes. However, they pose difficulties in smoothing the mesh nodes with varying degrees and require data augmentation to address the node input sequence problem. Additionally, the required labeled high-quality meshes further limit the applicability of the proposed method. In this paper, we present graph-based smoothing mesh net (GMSNet), a lightweight neural network model for intelligent mesh smoothing. GMSNet adopts graph neural networks (GNNs) to extract features of the node’s neighbors and outputs the optimal node position. During smoothing, we also introduce a fault-tolerance mechanism to prevent GMSNet from generating negative volume elements. With a lightweight model, GMSNet can effectively smooth mesh nodes with varying degrees and remain unaffected by the order of input data. A novel loss function, MetricLoss, is developed to eliminate the need for high-quality meshes, which provides stable and rapid convergence during training. We compare GMSNet with commonly used mesh-smoothing methods on two-dimensional (2D) triangle meshes. Experimental results show that GMSNet achieves outstanding mesh-smoothing performances with 5% of the model parameters compared to the previous model, but offers a speedup of 13.56 times over the optimization-based smoothing.

Graph processing underpins a vast array of data-centric applications, serving as a crucial component in fields such as social network analysis, recommendation systems, bio-informatics, and search engines. As graph data grows in scale and complexity, high-performance graph processing is increasingly essential. Many graph processing tasks depend on efficient data structures to manage the sparsity typical of real-world graphs, where most vertices have limited connectivity. This sparsity poses challenges for memory and computational efficiency in large-scale graph processing, and conventional sparse formats like Compressed Sparse Row (CSR) often struggle with memory and computation inefficiencies when handling massive graphs. To address these challenges, we introduce GraphCSR, a degree-equalized CSR format specifically tailored to enhance the spatio-temporal efficiency of distributed graph processing across various tasks. GraphCSR aggregates low-degree vertices into synthetic high-degree ones and applies group-wise compression to reduce storage overhead by recording only the starting index for each aggregated group. This reduces memory usage and supports batch-memory access to improve performance. Our extensive evaluations in various graph processing algorithms and datasets demonstrate that GraphCSR not only reduces the memory footprint required for large-scale graphs, but also improves performance across multiple types of graph processing tasks, outperforming popular sparse storage formats. Furthermore, when deployed on a production-scale supercomputer with 79,024 nodes, GraphCSR achieved a graph processing throughput that exceeded the top-ranked system on the Graph500 benchmark.

Sparse iterative solvers are commonly used in various fields. However, certain essential kernels of these solvers, such as sparse triangular solves (SpTRSV), present significant challenges for efficient parallelization due to data dependencies . Previous methods, like level-scheduling or multi-coloring, typically involve creating a Task Dependency Graph (TDG) to represent data dependencies and identify independent sets from the TDG for parallel execution. However, these approaches often result in limited parallelism with substantial synchronization overheads or negatively impact the solver convergence rate. This article introduces DCSolver , a Divide-and-Conquer (DC) framework designed to efficiently parallelize sparse solvers with data dependencies on GPUs. To achieve this, we break down the solver TDG into independent subgraphs, allowing us to exploit both coarse-grained and fine-grained parallelism. To efficiently allocate GPU threads for subgraphs with varying degrees of parallelism, we have developed an adaptive in-warp scheduling strategy. Additionally, we propose a hybrid parallelization scheme in DCSolver, which involves employing different parallel approaches for different DC recursions to achieve a more optimal balance between parallelism and convergence for solvers. To evaluate the effectiveness of DCSolver, we apply it to two preconditioned Krylov subspace solvers and an unstructured mesh Computational Fluid Dynamics (CFD) solver. Our results show that when compared with the state-of-the-art methods, DCSolver accelerates the time-to-solution of solvers by an average speedup of up to 26.19X.

Bipartite graph clustering (BGC) has emerged as a fast-growing research in the clustering community. Despite BGC has achieved promising scalability, most variants still suffer from the following concerns: a) Susceptibility to noisy features. They construct bipartite graphs in the raw feature space, inducing poor robustness to noisy features. b) Inflexible anchor selection strategies. They usually select anchors through heuristic sampling or constrained learning methods, degrading flexibility. c) Partial structure mining. Existing methods are mainly built upon Linear Reconstruction Paradigm (LRP) from subspace clustering or Locally Linear Paradigm (LLP) from manifold learning, which partially exploit linear or locally linear structures, lacking a unified perspective to integrate global complementary structures. To this end, we propose a novel model, termedJoint Robust Embedding and Structural FusionBipartiteGraphClustering (JetBGC), which focuses on three aspects, namely robustness, flexibility, and complementarity. Concretely, we first introduce a robust embedding learning module to extract latent representation that can reduce the impact of noisy features. Then, we optimize anchors via a constraint-free strategy that can flexibly capture data distribution. Furthermore, we revisit the consistency and specificity of LRP and LLP, and design a new unified structural fusion strategy to integrate both linear and locally linear structures from a global perspective. Therefore, JetBGC unifies robust representation learning, flexible anchor optimization, and structural bipartite graph fusion in a framework. Extensive experiments on synthetic and real-world datasets validate our effectiveness against existing baselines.

It is widely accepted that graph representations such as the Compressed Sparse Row (CSR) format, directly affect the space and time complexities of graph processing. However, the standard CSR and its current variations are prone to high memory footprint and complicated calculations, which necessitates the development of more efficient graph processing techniques to save memory and reduce calculations. This paper presents SuperCSR, a more space-time-efficient CSR representation for fast graph processing. SuperCSR’s key idea is to leverage the law of sorted graphs, which would directly access adjacent vertex sets from an active vertex ID without complex indexing calculations and with a lower memory footprint.

Sparse Matrix-Vector Multiplication (SpMV) plays a crucial role in scientific computing, but severe load imbalance among threads restricts its performance. Previous load-balancing methods have primarily ignored the CPU’s cache line-based memory access characteristics and the impact of data locality during workload evaluation and partitioning, leading to limited effect in load balancing.

We propose the SaSpGEMM: a parallel sparse general matrix-matrix multiplication (SpGEMM) to avoid the overhead of sorting. The typical workflow of SpGEMM contains: size prediction, memory allocation, numeric calculation and sorting. However, sorting has always been overlooked as a bottleneck in the performance of SpGEMM. It constitutes an average of 30% in HASH and 55% in ESC during the calculation stage. The key idea behind SaSpGEMM is to leverage the compressed sparse row (CSR) storage format’s feature of increasing index order of elements in the same row which is sorted during preprocessing, preserving intermediate products ordered consistently. To achieve this, we introduce a linked list-based accumulator (LLA) designed for batch insertion while maintaining order with a low time complexity. We provide a comprehensive empirical evidence showing that SaSpGEMM outperforms other methods based on time complexity analysis. Compared to three state-of-the-art methods ESC, SPA, and HASH on both x86 (Intel Xeon Gold 6348) and ARM (Phytium2000+) architectures, our method achieves an average speedup of 2.82x, 5.24x, 1.16x (with a maximum speedup of 34.4x, 195x, 23.3x) on the Intel Xeon Gold 6348. On the Phytium2000+, it achieves an average speedup of 2.21x, 4,65x, 1.05x (with a maximum speedup of 40.93x, 146.7x, 9.03x).

Irregular-shaped General Matrix Multiplication (GEMM) is extensively used in diverse workloads, such as scientific simulation and deep learning. In response to energy efficiency constraints, low-power multi-core digital signal processors (DSPs) have emerged as a viable alternative architecture in HPC systems. This study examines the performance of existing GEMM implementations working on irregular-shaped matrices and observes sub-optimal outcomes due to deficiencies in memory optimization and core-level parallelism extraction. For multi-core DSPs in FT-M7032, a CPU-DSP heterogeneous processor for HPC, we introduce dspIMM - a new multi-core parallel implementation for irregular-shaped matrix multiplications. dspIMM incorporates a new loop ordering, stack-space optimization, multi-dimensional core-level parallelization, a communication-computation overlap implementation with data prefetch across loops, and blocking optimization. These optimizations effectively enhance memory access and the core-level parallelism in irregular-shaped GEMMs. Experimental results demonstrate that the proposed communication-computation overlap optimization achieves the highest performance improvement in dspIMM, and the average speedup of dspIMM over previous implementations finally achieves up to 3.34 times.

Sparse matrix reordering is an important step in Cholesky decomposition. By reordering the rows and columns of the matrix, the time of computation and storage cost can be greatly reduced. With the proposal of various reordering algorithms, the selection of suitable reordering methods for various matrices has become an important research topic. In this paper, we propose a method to predict the optimal reordering method by visualizing sparse matrices in chunks in a parallel manner and feeding them into a deep convolutional neural network. The results show that the theoretical performance can reach 95% of the optimal performance, the prediction accuracy of the method can reach up to 85%, the parallel framework achieves an average speedup ratio of 11.35 times over the serial framework, and the performance is greatly improved compared with the traversal selection method on large sparse matrices.

Large parallel applications have heavily used MPI (Massage Passing Interface) collectives that support portable and efficient group communication operations. MPI_Bcast is one of the most commonly used MPI collectives that broadcast data to all processes of the communication domain. However, traditional software-based broadcast algorithms fail to fully utilize modern interconnection networks’ advanced features such as offloading collectives to the network hardware for efficient group communications. Besides, the semantic gap between MPI_Bcast and hardware multicast of underlying interconnects presents challenges for offload-based algorithms to accelerate MPI_Bcast for a wide range of message sizes.In this paper, we propose a hardware-software co-design MPI_Bcast by efficiently leveraging the NIC-based collective offload provided by Tianhe-express interconnect, which completely precludes the involvement of CPU to accelerate message broadcast. We detail this broadcast mechanism that can be adaptively tuned to offload MPI_Bcast operations from the CPU to the NIC for various message and system sizes. In addition, we further propose a topology-aware broadcast design in conjunction with this offload method to significantly reduce the broadcast latency by constructing the optimal global inter-node communication tree. We implement and evaluate the proposed Tianhe-Express Offload-based Broadcast (TOB) design on Tianhe-2A and Tianhe-EP supercomputers. Extensive experiments have been conducted to evaluate TOB performance at both microbenchmark and application levels. Our solution offers up to 4.94x significant performance speedup at the microbenchmark level over state-of-the-art MPI libraries. For the application-level evaluation, our technique accelerates scientific applications by a maximum speedup of 1.34x.

Restoring and colorizing old films is a comprehensive video enhancement task, marked by the presence of heterogeneous and structured degradations. Our DeepEnhancer addresses this challenge through a unified pipeline that combines restoration and colorization. In this workflow, we implement a bidirectional propagation strategy. Specifically, we incorporate second-order feature alignment to reduce the accumulation of inaccuracies in optical flow estimation. Simultaneously, we utilize cross-scale long-term attention mechanisms to model correlations within hidden states, thereby ensuring spatial and temporal consistency. To address the notable content loss in aging films, we introduce a temporally consistent focal transformer guided by global information. This transformer utilizes various window levels with distinct sub-window sizes to seamlessly integrate fine-grained and coarse-grained features. Comprehensive experimental results conclusively demonstrate the superiority of our model in both quantitative and qualitative comparisons when compared to existing approaches.

Due to data conflicts or data dependences, exploiting shared memory parallelism on unstructured mesh applications is highly challenging. The prior approaches are neither general nor scalable on emerging many-core processors. This paper presents a general and scalable shared memory approach for unstructured mesh computations. We recursively divide and reorder an unstructured mesh to construct a task dependency tree (TDT), where massive parallelism is exposed and data conflicts as well as data dependences are respected. We propose two recursion strategies to support popular programming models on both CPUs and GPUs for TDT. We evaluate our approach by applying it to an industrial unstructured Computational Fluid Dynamics (CFD) software. Experimental results show that our approach significantly outperforms the prior shared memory approaches, delivering up to 8.1× performance improvement over the engineer-tuned implementations.

Exploiting matrix symmetry to halve memory footprint offers an opportunity for accelerating memory-bound computations like Sparse Matrix-Vector Multiplication (SpMV). However, symmetric SpMV incurs data conflicts when concurrently writing the output vector. Previous approaches fail to address this issue efficiently. This paper proposes DCS-SpMV, a Divide-and-Conquer (DC) algorithm for efficient Symmetric SpMV. The key idea is to recursively divide the matrix-induced conflict graph into independent subgraphs for parallel execution, and construct separate subgraphs to avoid data conflicts. Our DC algorithm transforms the input matrix into a low-conflict part and a high-conflict part, which motivates us to design a conflict-aware hybrid solution that executes these two parts using DCS-SpMV and traditional SpMV respectively. We develop a machine learning model to predict an optimal hybrid implementation for a given matrix and architecture. We evaluate our work on both X86 and ARM CPUs, demonstrating significant performance improvement over the state-of-the-art.

In the design process of nuclear reactors, high-fidelity numerical simulations of thermo-hydraulic processes play a crucial role in evaluating the feasibility and safety of design schemes. These simulations involve complex physical processes such as coolant flow, thermal conduction, and heat transfer, necessitating the use of high-precision numerical algorithms and efficient parallel optimization methods. While mainstream commercial software is widely used in reactor design, it often falls short when it comes to meeting customized experimental scenarios and faces limitations in parallel scalability and physical model adaptability. Hence, the development of independent thermo-hydraulic parallel applications becomes essential. This paper focuses on conducting large-scale refined thermo-hydraulic analysis of core fuel rod bundles using the self-developed software YHACT. The paper elucidates the model algorithm design, parallel implementation, and analysis of test results. Two typical cases were selected for verification by comparing the obtained results with those from FLUENT. Emphasis was placed on the large-scale parallel simulation of fluid-solid coupled convection heat transfer phenomena between fuel rod bundles, utilizing a mesh scale exceeding 100 million cells. This breakthrough overcomes the scale limitations encountered with commercial software, and promises to foster further research and development in related fields.

Abstract Latent diffusion models (LDMs) have demonstrated remarkable success in generative modeling. It is promising to leverage the potential of diffusion priors to enhance performance in image and video tasks. However, applying LDMs to video super‐resolution (VSR) presents significant challenges due to the high demands for realistic details and temporal consistency in generated videos, exacerbated by the inherent stochasticity in the diffusion process. In this work, we propose a novel diffusion‐based framework, Temporal‐awareness Latent Diffusion Model (TempDiff), specifically designed for real‐world video super‐resolution, where degradations are diverse and complex. TempDiff harnesses the powerful generative prior of a pre‐trained diffusion model and enhances temporal awareness through the following mechanisms: 1) Incorporating temporal layers into the denoising U‐Net and VAE‐Decoder, and fine‐tuning these added modules to maintain temporal coherency; 2) Estimating optical flow guidance using a pre‐trained flow net for latent optimization and propagation across video sequences, ensuring overall stability in the generated high‐quality video. Extensive experiments demonstrate that TempDiff achieves compelling results, outperforming state‐of‐the‐art methods on both synthetic and real‐world VSR benchmark datasets. Code will be available at https://github.com/jiangqin567/TempDiff

Summary The development of computational fluid dynamics (CFD) highly depends on high‐performance computers. Computer hardware has evolved rapidly, yet scalable CFD parallel software remains scarce. In this article, we design a highly scalable CFD parallel paradigm for both homogeneous and heterogeneous supercomputers. The paradigm achieves the separation of communication and computation and automatically adapts to various solvers and hardware environments, thus reducing programming difficulties and increasing automatic parallelization. Meanwhile, the number of communications is greatly reduced and the scalability of the program is improved through implementing centralized communication and two‐level partitioning techniques. Complex flow problems for real aircraft were then computed on different hardware platforms with a grid size of ten billion. The homogeneous computer hardware includes Intel Xeon Gold 6258R and Phytium 2000+ processors, and the heterogeneous computer platforms include NVIDIA Tesla V100 and SW26010 processors. High parallel efficiency was obtained on all computer platforms, verifying that the paradigm has good automatic parallelization, scalability, and stability. The paradigm in this article has an important reference value for CFD massively parallel computing and can promote the development and application of CFD technology.

Unsupervised multi-view bipartite graph clustering (MVBGC) is a fast-growing research, due to promising scalability in large-scale tasks. Although many variants are proposed by various strategies, a common design is to construct the bipartite graph directly from the input data, i.e. only consider the unidirectional “encoding” process. However, “encoding-decoding” mechanism is a popular design for deep learning, the most representative one is auto-encoder (AE). Enlightened by this, this paper rethinks existing MVBGC paradigms and transfers the “encoding-decoding” design into graph machine learning, and proposes a novel framework termed auto-encoding multi-view bipartite graph clustering (BGAE), which integrates encoding, bipartite graph construction, and decoding modules in a self-supervised learning manner. The encoding module extracts a latent joint representation from the input data, the bipartite graph construction module learns a bipartite graph with connectivity constraint in latent semantic space, and the decoding module recreates the input data via the bipartite graph. Therefore, our novel BGAE combines representation learning, bipartite graph learning, reconstruction learning, and label inference into a unified framework. All the modules are seamlessly integrated and mutually reinforcing for clustering-friendly purposes. Extensive experiments verify the superiority of our novel design and the significance of “decoding” process. To the best of our knowledge, this is the first attempt to explore “encoding-decoding” design in traditional MVBGC. The code is provided athttps://github.com/liliangnudt/BGAE.

Textual adversarial attacks expose the vulnerabilities of text classifiers and can be used to improve their robustness. Previous context-aware attack models suffer from several limitations. They generally rely on out-of-date substitutes, solely consider the gold label probability, and use the greedy search when generating adversarial examples, often limiting the attack efficiency. To tackle these issues, we proposeMC-PDBS, aMoreContext-aware textual adversarial attack model usingProbabilityDifference-guidedBeamSearch. MC-PDBS generates substitutes using the newest perturbed text sequences in each attack iteration, enabling the generation of more context-aware adversarial examples. The probability difference is an overall consideration of the probabilities of all class labels, which is more effective than the gold label probability in guiding the selection of attack paths. In addition, the beam search enables MC-PDBS to search attack paths from multiple search channels, thereby avoiding the limited search space problem. Extensive experiments and human evaluation demonstrate that MC-PDBS outperforms previous best models in a series of evaluation metrics, particularly bringing up to a +19.5% attack success rate. Extensive analyses further confirm the effectiveness of MC-PDBS.

Knowledge distillation (KD) is a widely used approach to transfer knowledge from a cumbersome network (also known as a teacher) to a lightweight network (also known as a student). However, even though the accuracies of different teachers are similar, the fixed student's accuracies are significantly different. We find that teachers with more dispersed secondary soft probabilities are more qualified to play their roles. Therefore, an indicator, i.e., the standard deviation σ of secondary soft probabilities, is introduced to choose the teacher. Moreover, to make a teacher's secondary soft probabilities more dispersed, a novel method, dubbed pretraining the teacher under dual supervision (PTDS), is proposed to pretrain a teacher under dual supervision. In addition, we put forward an asymmetrical transformation function (ATF) to further enhance the dispersion degree of the pretrained teachers' secondary soft probabilities. The combination of PTDS and ATF is termed knowledge distillation with a customized teacher (KDCT). Extensive empirical experiments and analyses are conducted on three computer vision tasks, including image classification, transfer learning, and semantic segmentation, to substantiate the effectiveness of KDCT.

Unsupervised visual representation learning has gained much attention from the computer vision community because of the recent contrastive learning achievements. Current work mainly adopts instance discrimination as the pretext task, which treats every single instance as a different class (negative) and uses a collection of data augmentation techniques to generate more examples (positive) for each class. The idea is straightforward and efficient but will generally cause similar instances to be classified into different classes. Such problem has been defined as “class collision” in some previous works and is shown to hurt the representation ability. Motivated by this observation, we present a solution to address this issue by filtering similar negative examples from each mini-batch. Concretely, we model the problem as a Determinantal Point Process (DPP) so that similar instances can be filtered stochastically, and diverse samples are expected to be sampled for contrastive training. Besides, we further introduce a priority term for each instance, which indicates the hardness of its positives, so that instances with more hard positives are more likely to be sampled for contributing to the optimization. Our sampling can be efficiently implemented in a feed-forward manner and further accelerated by our encouraged complement DPP. Extensive experimental results demonstrate our priority over the standard setup of contrastive learning.

Continual Learning (CL) studies the problem of developing a robust model that can learn new tasks while retaining previously learned knowledge. However, the current CL methods exclusively focus on data with annotations, disregarding that unlabelled data is the mainstream in real-world applications. To close this research gap, this study concentrates on continual self-supervised learning, which is plagued by challenges of memory over-fitting and class imbalance. Besides, these challenges are exacerbated throughout incremental training. Aimed at addressing these challenges from both loss and data perspectives, we introduce a framework, Adaptive Self-supervised Continual Learning (ASCL). Specifically, we devise an Adaptive Sharpness-Aware Minimization (ASAM) module responsible for identifying flatter local minima in the loss landscape with a smaller memory over-fitting risk. Additionally, we design an Adaptive Memory Enhancement (AME) module responsible for rebalancing self-supervised loss with new and old tasks from a data perspective. Finally, the adaptive mechanisms in AME and ASAM modules dynamically adjust the loss landscape sharpness and memory enhancement strength with the feedback of intermediate training results. The results of our extensive experiments demonstrate the state-of-the-art performance of our methods in continual self-supervised learning scenarios across multiple datasets.

Global communication may affect the scalability of some parallel applications. Message Passing Interface (MPI) provides some commonly used collective communication Application Programming Interface(API). Allreduce is one of the APIs that is mostly used on parallel applications. Small message Allreduce is useful for dot products and solving linear systems. This paper proposes a medium and small message Allreduce for the Tianhe series system. For intra-node reduction/broadcast, this paper proposed a cache-aware tree and shared memory implementation. Besides, this paper proposes broadcast merge and cache line awareness methods to improve performance further. For inter-node communication, this paper proposes zero event Remote Direct Memory Access(RDMA) to avoid event overhead on the Tianhe system. In addition, a zero event immediate-data RDMA(Imm-RDMA) is used to optimize small message RDMA. Through experiments, for 16384 MPI processes, GLEX Allreduce achieves 2.4-5.1 times speedup compared to MPI. Compared to other collective communication libraries on Infiniband(IB) and Omni-Path, GLEX_Allreduce achieves similar or better performance.

Partial differential equations (PDEs) are an essential computational kernel in physics and engineering. With the advance of deep learning, physics-informed neural networks (PINNs), as a mesh-free method, have shown great potential for fast PDE solving in various applications. To address the issue of low accuracy and convergence problems of existing PINNs, we propose a self-training physics-informed neural network, ST-PINN. Specifically, ST-PINN introduces a pseudo label based self-learning algorithm during training. It employs governing equation as the pseudo-labeled evaluation index and selects the highest confidence examples from the sample points to attach the pseudo labels. To our best knowledge, we are the first to incorporate a self-training mechanism into physics-informed learning. We conduct experiments on five PDE problems in different fields and scenarios. The results demonstrate that the proposed method allows the network to learn more physical information and benefit convergence. The ST-PINN outperforms existing physics-informed neural network methods and improves the accuracy by a factor of 1.33x-2.54x.

4K Video Super-Resolution (VSR) presents a challenging task in video processing, as most existing VSR models have high computational complexity, limiting their application to high-resolution videos, particularly for 4K resolution videos. To address this issue, we propose a novel Guided Spatio-Temporal Video Super-Resolution network (GST-VSR) designed to perform 4K VSR on a single GPU. The proposed method comprises two key components: the Spatio-Temporal Alignment Network (STAN) and the Super-resolution Reconstruction Network (SRN), which work together to enhance the quality of the output frames. The STAN is responsible for extracting highly relevant features in frames and aligning the reference frame with the neighboring frames at the feature level to maintain temporal consistency. The SRN fuses high-quality features into the final high-resolution frames. Unlike existing methods, our proposed approach does not require explicit optical flow estimation, making it more efficient and less computationally demanding. To facilitate the training and testing of the compared models, we have established a new dataset, Pixabay-Set, consisting of 145 videos suitable for the 4K VSR task. Experimental results on the test dataset show that the proposed method achieves competitive performance compared to state-of-the-art models. In summary, our proposed GST-VSR network provides an effective solution to the challenging task of 4K VSR.

Message Passing Interface (MPI) programs often experience performance slowdowns due to collective communication operations, like broadcasting and reductions. As modern CPUs integrate more processor cores, running multiple MPI processes on shared-memory machines to take advantage of hardware parallelism is becoming increasingly common. In this context, it is crucial to optimize MPI collective communications for shared-memory execution. However, existing MPI collective implementations on shared-memory systems have two primary drawbacks. The first is extensive redundant data movements when performing reduction collectives, and the second is the ineffective use of non-temporal instructions to optimize streamed data processing. To address these limitations, this paper proposes two optimization techniques that minimize data movements and enhance the use of non-temporal instructions. We evaluated our techniques by integrating them into the OpenMPI library and tested their performance using micro-benchmarks and real-world applications running on two multi-core clusters. Experimental results show that our approach significantly outperforms existing techniques, yielding a 1.2--6.4x performance improvement.

Single-cell RNA-sequencing (scRNA-seq) has emerged as a powerful technique for studying gene expression patterns at the single-cell level. Inferring gene regulatory networks (GRNs) from scRNA-seq data provides insight into cellular phenotypes from the genomic level. However, the high sparsity, noise and dropout events inherent in scRNA-seq data present challenges for GRN inference. In recent years, the dramatic increase in data on experimentally validated transcription factors binding to DNA has made it possible to infer GRNs by supervised methods. In this study, we address the problem of GRN inference by framing it as a graph link prediction task. In this paper, we propose a novel framework called GNNLink, which leverages known GRNs to deduce the potential regulatory interdependencies between genes. First, we preprocess the raw scRNA-seq data. Then, we introduce a graph convolutional network-based interaction graph encoder to effectively refine gene features by capturing interdependencies between nodes in the network. Finally, the inference of GRN is obtained by performing matrix completion operation on node features. The features obtained from model training can be applied to downstream tasks such as measuring similarity and inferring causality between gene pairs. To evaluate the performance of GNNLink, we compare it with six existing GRN reconstruction methods using seven scRNA-seq datasets. These datasets encompass diverse ground truth networks, including functional interaction networks, Loss of Function/Gain of Function data, non-specific ChIP-seq data and cell-type-specific ChIP-seq data. Our experimental results demonstrate that GNNLink achieves comparable or superior performance across these datasets, showcasing its robustness and accuracy. Furthermore, we observe consistent performance across datasets of varying scales. For reproducibility, we provide the data and source code of GNNLink on our GitHub repository: https://github.com/sdesignates/GNNLink.

Abstract The development of a basic scalable preprocessing tool is the key routine to accelerate the entire computational fluid dynamics (CFD) workflow toward the exascale computing era. In this work, a parallel preprocessing tool, called ParTransgrid, is developed to translate the general grid format like CFD General Notation System into an efficient distributed mesh data format for large‐scale parallel computing. Through ParTransgrid, a flexible face‐based parallel unstructured mesh data structure designed in Hierarchical Data Format can be obtained to support various cell‐centered unstructured CFD solvers. The whole parallel preprocessing operations include parallel grid I/O, parallel mesh partition, and parallel mesh migration, which are linked together to resolve the run‐time and memory consumption bottlenecks for increasingly large grid size problems. An inverted index search strategy combined with a multi‐master‐slave communication paradigm is proposed to improve the pairwise face matching efficiency and reduce the communication overhead when constructing the distributed sparse graph in the phase of parallel mesh partition. And we present a simplified owner update rule to fast the procedure of raw partition boundaries migration and the building of shared faces/nodes communication mapping list between new sub‐meshes with an order of magnitude of speed‐up. Experiment results reveal that ParTransgrid can be easily scaled to billion‐level grid CFD applications, the preparation time for parallel computing with hundreds of thousands of cores is reduced to a few minutes.

For a class of uncertain nonlinear systems with actuator failures, the event-triggered prescribed settling time consensus adaptive compensation control method is proposed. The unknown form of actuator failures may occur in practical applications, resulting in system instability or even control failure. In order to effectively deal with the above problems, a neural network adaptive control method is developed to ensure that the system states rapidly converge in the event of failure and compensate for the failures of actuator. Meanwhile, a nonlinear transformation function is introduced to make sure that the tracking error converges for the predefined interval within a prescribed settling time, which makes that the convergence time can be preset. Furthermore, a finite-time event-triggered compensation control strategy is established by the backstepping technology. Under this strategy, the system not only can rapidly stabilize in finite time but also can effectively save network bandwidth. In addition, the states of the system are globally uniformly bounded. Finally, the theoretical analysis and simulation experiments validate the effectiveness of the proposed method.

General Matrix Multiplication (GEMM) has a wide range of applications in scientific simulation and artificial intelligence. Although traditional libraries can achieve high performance on large regular-shaped G EMMs, they often behave not well on irregular-shaped G EMMs, which are often found in new algorithms and applications of high-performance computing (HPC). Due to energy efficiency constraints, low-power multi-core digital signal processors (DSPs) have become an alternative architecture in HPC systems. Targeting multi-core DSPs in FT-m7032, a prototype CPU-DSPs heterogeneous processor for HPC, an efficient implementation-ftIMM - for three types of irregular-shaped GEMMs is proposed. FtIMM supports automatic generation of assembly micro-kernels, two parallelization strategies, and auto-tuning of block sizes and parallelization strategies. The experiments show that ftIMM can get better performance than the traditional GEMM implementations on multi-core DSPs in FT-m7032, yielding on up to 7.2x performance improvement, when performing on irregular-shaped GEMMs. And ftIMM on multi-core DSPs can also far outperform the open source library on multi-core CPUs in FT-m7032, delivering up to 3.1 x higher efficiency.

Graph algorithms, such as Breadth First Search (BFS), Single Source Shortest Path (SSSP), PageRank (PR), and Connected Components (CC), are increasingly important in big data processing and analytics. As graph scales (numbers of vertices and edges) have increased from billions to trillions, Supercomputers have huge numbers (up to hundreds of thousands) of computing nodes (CNs) that can provide ultra-high aggregate computing power and memory capacity, thus being particularly suitable for processing extreme-scale graphs with trillions of vertices and edges. However, existing cluster-based graph-parallel systems perform poorly when deployed on supercomputers, since their partitioning methods overlook the hierarchical nature of supercomputer networks and incur prohibitive communication storm. This paper presents XTree, an efficient traversal-based partitioning method for minimizing communication overhead of graph processing on supercomputers. We observe that supercomputers' huge numbers of CNs are usually organized into hierarchical communication domains, which can be modeled as a domain tree where communication in lower-level domains is significantly faster than that in higher-level ones. Therefore, the key idea of XTree's partitioning is to exploit hierarchical locality by viewing the graph as a BFS tree and leveraging the topology knowledge to map the graph's BFS tree onto the domain tree, We evaluate the effectiveness of XTree by running various graph algorithms, on both real-world big graphs and synthetic trillion-scale graphs. XTree substantially reduces communication overhead and achieves orders of magnitude speedup against the Graph500 reference implementations with the state-of-the-art 2D-decomposition partitioning.

BACKGROUND: Building biological networks with a certain function is a challenge in systems biology. For the functionality of small (less than ten nodes) biological networks, most methods are implemented by exhausting all possible network topological spaces. This exhaustive approach is difficult to scale to large-scale biological networks. And regulatory relationships are complex and often nonlinear or non-monotonic, which makes inference using linear models challenging. RESULTS: In this paper, we propose a multi-layer perceptron-based differential equation method, which operates by training a fully connected neural network (NN) to simulate the transcription rate of genes in traditional differential equations. We verify whether the regulatory network constructed by the NN method can continue to achieve the expected biological function by verifying the degree of overlap between the regulatory network discovered by NN and the regulatory network constructed by the Hill function. And we validate our approach by adapting to noise signals, regulator knockout, and constructing large-scale gene regulatory networks using link-knockout techniques. We apply a real dataset (the mesoderm inducer Xenopus Brachyury expression) to construct the core topology of the gene regulatory network and find that Xbra is only strongly expressed at moderate levels of activin signaling. CONCLUSION: We have demonstrated from the results that this method has the ability to identify the underlying network topology and functional mechanisms, and can also be applied to larger and more complex gene network topologies.

3-D magnetotelluric (MT) forward modeling has always been faced with the problems of high memory requirements and long computing time. In this article, we design a scalable parallel algorithm for 3-D MT finite element modeling in anisotropic media. The parallel algorithm is based on the distributed mesh storage, including multiple parallel granularities, and is implemented through multiple tools. Message-passing interface (MPI) is used to exploit process parallelisms for subdomains, frequencies, and solving equations. Thread parallelisms for merge sorting, element analysis, matrix assembly, and imposing Dirichlet boundary conditions are developed by Open Multi-Processing (OpenMP). We validate the algorithm through several model simulations and study the effects of topography and conductivity anisotropy on apparent resistivities and phase responses. Scalability tests are performed on the Tianhe-2 supercomputer to analyze the parallel performance of different parallel granularities. Three parallel direct solvers Supernodal LU (SUPERLU), MUltifrontal Massively Parallel sparse direct Solver (MUMPS), and Parallel Sparse matriX package (PASTIX) are compared in solving sparse systems of equations. As a result, reasonable parallel parameters are suggested for practical applications. The developed parallel algorithm is proven to be efficient and scalable.

As the era of exascale supercomputing is coming, it is vital for next-generation supercomputers to find appropriate applications with high social and economic benefit. In recent years, it has been widely accepted that extremely-large graph computation is a promising killer application for supercomputing. Although Tianhe series supercomputers are leading in the world-wide competition of supercomputing (ranked No. 1 in the Top500 list for six times), previously they had been inefficient in graph computation according to the Graph500 list. This is mainly because the previous graph processing system cannot leverage the advanced hardware features of Tianhe supercomputers. To address the problem, in this paper we present our integrated optimizations for improving the graph computation performance on our next-generation Tianhe supercomputing system, mainly including sorting with buffering for heavy vertices, vectorized searching with SVE (Scalable Vector Extension) on matrix2000+ CPUs, and group communication on the proprietary interconnection network. Performance evaluation on a subset of the Tianhe supercomputer (with 512 nodes and 196,608 cores) shows that our customized graph processing system effectively improves the graph search performance and achieves the BFS performance of 2131.98 GTEPS.

In this paper, we propose some efficient data redistribution algorithms for redistributing matrices from 1D or 2D irregular format to block cyclic data distribution (BCDD) format, which can be much faster than the BLACS routinePXGEMR2D. These algorithms can be used to combine direct methods with iterative methods. The proposed algorithms divide the communication into two phases: one for processes in the same column and the other for processes in the same row, and the whole data redistribution task is divided into several independent sub-communications. The communication time can be reduced a lot compared with BLACS. Performance results show that our algorithms can be$2\times$–$5\times$faster than the BLACS routinePXGEMR2Dwhen using 4096 processes and the experiments are performed on Tianhe-2A supercomputer.

This paper endeavors to leverage spatio-temporal visual cues to improve video-based action detection. As a result, a NOn-Local Action detector based on anchor-free called NOLA is proposed, which is built off a recent moving center detector (MOC) and further extends it by efficiently aggregating long-range spatio-temporal information. In detail, a significantly efficient spatio-temporal motion-aware non-local block is explored to provide global motion contexts for the entire predictive branches of MOC. This byproduct can make the large batch samples run on a resource limited device. Besides, a light-weighted data augmentation method termed clip augmentation designed for video-based tasks is proposed, which serves to improve the generalization ability of the detector with economical scale-and-addition operation. NOLA works with two above schemes in real-time as well. Experiments on two benchmark datasets show that NOLA significantly exceeds MOC. Compared to other existing methods,, NOLA reaches the state-of-the-art, in terms of video-level mean of average precision (video mAP).

Convolutional Neural Networks (CNNs) are a class of deep neural networks. The image to column (im2col) procedure is an important step for CNN and consumes about 28.8% of the whole inference time. In this paper, we present an efficient im2col algorithm, name im2cole (word “e” means efficient). The condition with different stride and pad in im2cole is well handled and the judgements in the innermost loop are removed. The procedure with pad = 1 is split into three conditions. This will reduce the pause of CPU instruction pipeline. The performances of the presented im2cole algorithm are reported with different inputs. Some discussion and performance issues are also reported. The experimental results show that the overall performance speedup of im2cole ranges from 2.12 to 4.33 compared with the original algorithm. The real application with Darknet shows that im2cole can get 20.75% whole performance improvement.

Compressive Sensing for Magnetic Resonance Imaging (CS-MRI) aims to reconstruct Magnetic Resonance (MR) images from under-sampled raw data. There are two challenges to improve CS-MRI methods, i.e. designing an under-sampling algorithm to achieve optimal sampling, as well as designing fast and small deep neural networks to obtain reconstructed MR images with superior quality. To improve the reconstruction quality of MR images, we propose a novel deep convolutional neural network architecture for CS-MRI named MRCSNet. The MRCSNet consists of three sub-networks, a compressive sensing sampling sub-network, an initial reconstruction sub-network, and a refined reconstruction sub-network. Experimental results demonstrate that MRCSNet generates high-quality reconstructed MR images at various under-sampling ratios, and also meets the requirements of real-time CS-MRI applications. Compared to state-of-the-art CS-MRI approaches, MRCSNet offers a significant improvement in reconstruction accuracies, such as Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity (SSIM). Besides, it reduces the reconstruction error evaluated by the Normalized Root-Mean-Square Error (NRMSE). The source codes are available at https://github.com/TaihuLight/MRCSNet .

Nowadays, cloud co-location is often used for data centers to improve the utilization of computing resources. However, batch jobs in a Co-location Datacenter (CLD) are vulnerable to failures due to the competition for limited resources with online service jobs. Such failed batch jobs would be rescheduled and failed repeatedly, resulting in the waste of computing resources and instability of the computing clusters. Therefore, we propose a method to accurately predict the potential failures of batch jobs for CLD. The core of the proposed method is STLF (SMOTE Tomek and LightGBM [5] Framework), which is divided into three parts. First, we use the co-feature extraction method to generate Co-located Feature Dataset (CLFD). Then SMOTE Tomek is used to oversampling the CLFD to ensure that the classifier can learn more minority features. Finally, we use LightGBM classifier to predict batch jobs' failure. The performance experiments conducted on the Ali Trace 2018 dataset show that our proposed STLF significantly outperforms the existing popular classifiers in terms of the ROC curve, the area under the ROC curve (AUC), precision, and recall.

Nowadays, both online shopping and video sharing have grown exponentially. Although internet celebrities in videos are ideal exhibition for fashion corporations to sell their products, audiences do not always know where to buy fashion products in videos, which is a cross-domain problem called video-to-shop. In this paper, we propose a novel deep neural network, called Detect, Pick, and Retrieval Network (DPRNet), to break the gap between fashion products from videos and audiences. For the video side, we have modified the traditional object detector, which automatically picks out the best object proposals for every commodity in videos without duplication, to promote the performance of the video-to-shop task. For the fashion retrieval side, a simple but effective multi-task loss network obtains new state-of-the-art results on DeepFashion. Extensive experiments conducted on a new large-scale cross-domain video-to-shop dataset shows that DPRNet is efficient and outperforms the state-of-the-art methods on video-to-shop task.

In e-commerce, a growing number of user-generated videos are used for product promotion. How to generate video descriptions that narrate the user-preferred product characteristics depicted in the video is vital for successful promoting. Traditional video captioning methods, which focus on routinely describing what exists and happens in a video, are not amenable for product-oriented video captioning. To address this problem, we propose a product-oriented video captioner framework, abbreviated as Poet. Poet firstly represents the videos as product-oriented spatial-temporal graphs. Then, based on the aspects of the video-associated product, we perform knowledge-enhanced spatial-temporal inference on those graphs for capturing the dynamic change of fine-grained product-part characteristics. The knowledge leveraging module in Poet differs from the traditional design by performing knowledge filtering and dynamic memory modeling. We show that Poet achieves consistent performance improvement over previous methods concerning generation quality, product aspects capturing, and lexical diversity. Experiments are performed on two product-oriented video captioning datasets, buyer-generated fashion video dataset (BFVD) and fan-generated fashion video dataset (FFVD), collected from Mobile Taobao. We will release the desensitized datasets to promote further investigations on both video captioning and general video analysis problems.

With the rapid increase of the size of applications and the complexity of the supercomputer architecture, topology-aware process mapping becomes increasingly important. High communication cost has become a dominant constraint of the performance of applications running on the supercomputer. To avoid a bad mapping strategy which can lead to terrible communication performance, we propose an optimized heuristic topology-aware mapping algorithm (OHTMA). The algorithm attempts to minimize the hop-byte metric that we use to measure the mapping results. OHTMA incorporates a new greedy heuristic method and pair-exchange-based optimization. It reduces the number of long-distance communications and effectively enhances the locality of the communication. Experimental results on the Tianhe-3 exascale supercomputer prototype indicate that OHTMA can significantly reduce the communication costs.

Molecular dynamics (MD) is a computer simulation method of studying physical movements of atoms and molecules that provide detailed microscopic sampling on molecular scale. With the continuous efforts and improvements, MD simulation gained popularity in materials science, biochemistry and biophysics with various application areas and expanding data scale. Assisted Model Building with Energy Refinement (AMBER) is one of the most widely used software packages for conducting MD simulations. However, the speed of AMBER MD simulations for system with millions of atoms in microsecond scale still need to be improved. In this paper, we propose a parallel acceleration strategy for AMBER on the Tianhe-2 supercomputer. The parallel optimization of AMBER is carried out on three different levels: fine grained OpenMP parallel on a single CPU, single node CPU/MIC parallel optimization and multi-node multi-MIC collaborated parallel acceleration. By the three levels of parallel acceleration strategy above, we achieved the highest speedup of 25-33 times compared with the original program.

Convolution neural networks (CNNs) have been extensively used in machine learning applications. The most time-consuming part of CNNs are convolution operations. A common approach to implementing convolution operations is to recast them as general matrix multiplication, known as the im2col+GEMM approach. There are two main drawbacks of this approach. One is that large additional memory space is required. The other is the packing on the input elements of convolution operations are not memory-efficient enough. In this paper, we present a new parallel convolution algorithm using implicit matrix multiplication on multi-core CPUs. In comparison with Im2col+GEMM, our new algorithm can reduce the memory footprints and improve the packing efficiency. The experiment results on two ARV8-based multi-core CPUs demonstrate that our new algorithm gives much better performance and scalability than the im2col+GEMM method in most cases.

Summary Sparse matrix‐vector multiplication (SpMV) is an essential kernel in sparse linear algebra and has been studied extensively on all modern processor and accelerator architectures. Compressed Sparse Row (CSR) is a frequently used format for sparse matrices storage. However, CSR‐based SpMV has poor performance on processors with vector units. In order to take full advantage of SIMD acceleration technology in SpMV, we proposed a new matrix storage format called CSR‐SIMD. The new storage format compresses the non‐zero elements into many variable‐length data fragments with consecutive memory access addresses. Thus, the data locality of sparse matrix A and dense vector x expands and the floating‐point operations for each fragment can be completely calculated by vectorized implementation on wide SIMD units. Our experimental results indicate that CSR‐SIMD has better storage efficiency and low‐overhead for format conversion. Besides, the new format achieves high scalability on wide SIMD units. In comparison with the CSR‐based and BCSR‐based SpMV, CSR‐SIMD obtains better performance on FT1500A, Intel Xeon, and Intel Xeon Phi.

Collaborative Filtering (CF), a well-known approach in producing recommender systems, has achieved wide use and excellent performance not only in research but also in industry. However, problems related to cold start and data sparsity have caused CF to attract an increasing amount of attention in efforts to solve these problems. Traditional approaches adopt side information to extract effective latent factors but still have some room for growth. Due to the strong characteristic of feature extraction in deep learning, many researchers have employed it with CF to extract effective representations and to enhance its performance in rating prediction. Based on this previous work, we propose a probabilistic model that combines a stacked denoising autoencoder and a convolutional neural network together with auxiliary side information (i.e, both from users and items) to extract users and items’ latent factors, respectively. Extensive experiments for four datasets demonstrate that our proposed model outperforms other traditional approaches and deep learning models making it state of the art.

Molecular dynamics (MD) is a computer simulation method of studying physical movements of atoms and molecules that provide detailed microscopic sampling on molecular scale. With the continuous efforts and improvements, MD simulation gained popularity in materials science, biochemistry and biophysics with various application areas and expanding data scale. Assisted Model Building with Energy Refinement (AMBER) is one of the most widely used software packages for conducting MD simulations. However, the speed of AMBER MD simulations for system with millions of atoms in microsecond scale still need to be improved. In this paper, we propose a parallel acceleration strategy for AMBER on Tianhe-2 supercomputer. The parallel optimization of AMBER is carried out on three different levels: fine grained OpenMP parallel on a single MIC, single-node CPU/MIC collaborated parallel optimization and multi-node multi-MIC collaborated parallel acceleration. By the three levels of parallel acceleration strategy above, we achieved the highest speedup of 25-33 times compared with the original program. Source Code: https://github.com/tianhe2/mAMBER.

Whole genome re-sequencing plays a crucial role in biomedical studies. The emergence of genomic big data calls for an enormous amount of computing power. However, current computational methods are inefficient in utilizing available computational resources. In this paper, we address this challenge by optimizing the utilization of the fastest supercomputer in the world - TH-2 supercomputer. TH-2 is featured by its neo-heterogeneous architecture, in which each compute node is equipped with 2 Intel Xeon CPUs and 3 Intel Xeon Phi coprocessors. The heterogeneity and the massive amount of data to be processed pose great challenges for the deployment of the genome analysis software pipeline on TH-2. Runtime profiling shows that SOAP3-dp and SOAPsnp are the most time-consuming components (up to 70% of total runtime) in a typical genome-analyzing pipeline. To optimize the whole pipeline, we first devise a number of parallel and optimization strategies for SOAP3-dp and SOAPsnp, respectively targeting each node to fully utilize all sorts of hardware resources provided both by CPU and MIC. We also employ a few scaling methods to reduce communication between different nodes. We then scaled up our method on TH-2. With 8192 nodes, the whole analyzing procedure took 8.37 hours to finish the analysis of a 300 TB dataset of whole genome sequences from 2,000 human beings, which can take as long as 8 months on a commodity server. The speedup is about 700x.

In this paper, we describe our experiment developing an implementation of the Linpack benchmark for TianHe-1, a petascale CPU/GPU supercomputer system, the largest GPU-accelerated system ever attempted before. An adaptive optimization framework is presented to balance the workload distribution across the GPUs and CPUs with the negligible runtime overhead, resulting in the better performance than the static or the training partitioning methods. The CPU-GPU communication overhead is effectively hidden by a software pipelining technique, which is particularly useful for large memory-bound applications. Combined with other traditional optimizations, the Linpack we optimized using the adaptive optimization framework achieved 196.7 GFLOPS on a single compute element of TianHe-1. This result is 70.1% of the peak compute capability and 3.3 times faster than the result using the vendor's library. On the full configuration of TianHe-1 our optimizations resulted in a Linpack performance of 0.563PFLOPS, which made TianHe-1 the 5th fastest supercomputer on the Top500 list released in November 2009.