Zhe Dai
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Zhe Dai has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Jian Zhang
China Aerodynamics Research and Development Center
3 shared publications - Jie Liu
Laboratory of Digitizing Software for Frontier Equipment, National University of Defense Technology, Changsha, Hunan, China
3 shared publications - Liang Deng
China Aerodynamics Research and Development Center
3 shared publications - Chuanfu Xu
National University of Defense Technology
2 shared publications - Haozhong Qiu
National University of Defense Technology College of Computer Science and Technology
2 shared publications - Jianbin Fang
College of Computer Science and Technology, National University of Defense Technology, Changsha, China
2 shared publications
Latest Publications
- Open Access
- 22 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 23 (1) , 1-24
HiDAP-BILU: Hierarchical Dependency-Aware Parallelism for Block ILU Preconditioner on GPUs
The block incomplete LU (BILU(0)) preconditioner is widely adopted for solving large-scale block-sparse linear systems arising from coupled partial differential equations (PDEs). However, strong inherent data dependencies and high memory bandwidth requirements of block matrix operations in the preconditioner pose significant challenges for efficient implementation on GPUs. Existing methods face the tradeoff between parallelism and convergence, and efforts to leverage the block properties remain limited. In this work, we introduce HiDAP-BILU for dependency-aware BILU(0) preconditioner on GPUs. HiDAP-BILU employs a block-centric hierarchical parallelism strategy to maximize concurrency. A block-to-block waiting scheme is proposed to preserve inter-block dependencies, while warp-level data distribution operations are utilized to maintain intra-block dependencies. Aligned with block-centric hierarchical parallelism, various architecture-aware optimizations are designed to minimize warp divergence and ensure coalesced memory access. Experiment results demonstrate that HiDAP-BILU achieves up to 4.77x and 3.10x speedups compared with state-of-the-art general methods for factorization and triangular solve, respectively. Additionally, it provides an average speedup of 3.11x in end-to-end BILU-preconditioned iterative solvers.
- 22 March 2026
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 23 (1) , 1-24
Open AccessHiDAP-BILU: Hierarchical Dependency-Aware Parallelism for Block ILU Preconditioner on GPUs
The block incomplete LU (BILU(0)) preconditioner is widely adopted for solving large-scale block-sparse linear systems arising from coupled partial differential equations (PDEs). However, strong inherent data dependencies and high memory bandwidth requirements of block matrix operations in the preconditioner pose significant challenges for efficient implementation on GPUs. Existing methods face the tradeoff between parallelism and convergence, and efforts to leverage the block properties remain limited. In this work, we introduce HiDAP-BILU for dependency-aware BILU(0) preconditioner on GPUs. HiDAP-BILU employs a block-centric hierarchical parallelism strategy to maximize concurrency. A block-to-block waiting scheme is proposed to preserve inter-block dependencies, while warp-level data distribution operations are utilized to maintain intra-block dependencies. Aligned with block-centric hierarchical parallelism, various architecture-aware optimizations are designed to minimize warp divergence and ensure coalesced memory access. Experiment results demonstrate that HiDAP-BILU achieves up to 4.77x and 3.10x speedups compared with state-of-the-art general methods for factorization and triangular solve, respectively. Additionally, it provides an average speedup of 3.11x in end-to-end BILU-preconditioned iterative solvers.
- Open Access
- 17 February 2026
- Published by Elsevier in SSRN Electronic Journal
Parallel implicit methods for an Unstructured CFD Simulator on GPU
The advancement of GPU technology has revolutionized Computational Fluid Dynamics (CFD) by enabling efficient generation of high-precision flow field data. However, a critical challenge lies in reconciling the fast convergence of data-dependent implicit schemes with GPU parallelism, as existing methods struggle to strike a balance between computational efficiency and convergence accuracy. This study systematically evaluates four implicit approaches implemented in the industrial solver FlowStar, focusing on efficiency-convergence trade-offs stemming from serial dependencies that limit GPU utilization. To address these limitations, a balanced coloring and hyperplane method is proposed to optimize hardware resource allocation, while a unified evaluation framework quantifies performance across five unstructured grid configurations. Experimental results demonstrate that the balanced coloring and hyperplane implementations achieve comprehensive speedups of 2.33× and 1.15× against the conventional level-set parallelization. This work provides actionable insights for designing parallel CFD algorithms that harmonize parallelism with numerical stability.
- 17 February 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessParallel implicit methods for an Unstructured CFD Simulator on GPU
The advancement of GPU technology has revolutionized Computational Fluid Dynamics (CFD) by enabling efficient generation of high-precision flow field data. However, a critical challenge lies in reconciling the fast convergence of data-dependent implicit schemes with GPU parallelism, as existing methods struggle to strike a balance between computational efficiency and convergence accuracy. This study systematically evaluates four implicit approaches implemented in the industrial solver FlowStar, focusing on efficiency-convergence trade-offs stemming from serial dependencies that limit GPU utilization. To address these limitations, a balanced coloring and hyperplane method is proposed to optimize hardware resource allocation, while a unified evaluation framework quantifies performance across five unstructured grid configurations. Experimental results demonstrate that the balanced coloring and hyperplane implementations achieve comprehensive speedups of 2.33× and 1.15× against the conventional level-set parallelization. This work provides actionable insights for designing parallel CFD algorithms that harmonize parallelism with numerical stability.
- Open Access
- 19 September 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 22 (3) , 1-25
DCSolver: Accelerating Sparse Iterative Solvers via Divide-and-Conquer on GPUs
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.
- 19 September 2025
- Published by Association for Computing Machinery (ACM) in ACM Transactions on Architecture and Code Optimization
- Vol. 22 (3) , 1-25
Open AccessDCSolver: Accelerating Sparse Iterative Solvers via Divide-and-Conquer on GPUs
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.
- Open Access
- 20 February 2024
- Published by Association for Computing Machinery (ACM)
- p. 109-119
Towards Scalable Unstructured Mesh Computations on Shared Memory Many-Cores
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.
- 20 February 2024
- Published by Association for Computing Machinery (ACM)
- p. 109-119
Open AccessTowards Scalable Unstructured Mesh Computations on Shared Memory Many-Cores
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.
- Open Access
- 1 June 2008
- Published by Oxford University Press (OUP) in Monthly Notices of the Royal Astronomical Society
- Vol. 388 (3) , 1409-1418
Self-similar structure of magnetized advection-dominated accretion flows and convection-dominated accretion flows
- 1 June 2008
- Published by Oxford University Press (OUP) in Monthly Notices of the Royal Astronomical Society
- Vol. 388 (3) , 1409-1418
Open AccessSelf-similar structure of magnetized advection-dominated accretion flows and convection-dominated accretion flows
- 1 November 2004
- Published by American Astronomical Society in The Astrophysical Journal
- Vol. 615 (1) , 359-365
Jet Break Time–Flux Density Relationship and Constraints on Physical Parameters of Gamma‐Ray Burst Afterglows
We derive a relation between the flux density $F_{\nu,j}$ at the light-curve break of a gamma-ray burst (GRB) afterglow and the break time $t_{j}$. The break is due to the transition from the spherical-like to jet-like evolution of the afterglow, when the Lorentz factor of the jet equals the inverse of the initial half-opening angle, i.e., $\gamma=1/\theta_0$. We show that this relation indeed behaves as $F_{\nu,j}\propto t_{j}^{-p}$ among GRBs for the slow-cooling case, where $p$ is the power-law index of electron distribution. A statistical analysis of the optical jet breaks of nine GRBs gives $p=2.10\pm 0.21$, which is consistent with the shock acceleration theory. The value of $p$ derived in this way is different from the observed temporal index $\alpha_2$ ($F_{\nu}\propto t^{-\alpha_{2}}$) of the late-time light curve after $t_{j}$, which suffers several uncertainties from the unclear hydrodynamics of the sideways expansion and exhibits a large dispersion. Our results not only confirm that the remnants of GRBs are standard candles, but also provide the first evidence that the physical parameters of relativistic shocks are universal, with the favored values $\epsilon_{e}\sim 0.1$ and $\epsilon_{B}\sim 10^{-3}$.Comment: 11 pages including 3 color figure- 1 November 2004
- Published by American Astronomical Society in The Astrophysical Journal
- Vol. 615 (1) , 359-365
Jet Break Time–Flux Density Relationship and Constraints on Physical Parameters of Gamma‐Ray Burst Afterglows
We derive a relation between the flux density $F_{\nu,j}$ at the light-curve break of a gamma-ray burst (GRB) afterglow and the break time $t_{j}$. The break is due to the transition from the spherical-like to jet-like evolution of the afterglow, when the Lorentz factor of the jet equals the inverse of the initial half-opening angle, i.e., $\gamma=1/\theta_0$. We show that this relation indeed behaves as $F_{\nu,j}\propto t_{j}^{-p}$ among GRBs for the slow-cooling case, where $p$ is the power-law index of electron distribution. A statistical analysis of the optical jet breaks of nine GRBs gives $p=2.10\pm 0.21$, which is consistent with the shock acceleration theory. The value of $p$ derived in this way is different from the observed temporal index $\alpha_2$ ($F_{\nu}\propto t^{-\alpha_{2}}$) of the late-time light curve after $t_{j}$, which suffers several uncertainties from the unclear hydrodynamics of the sideways expansion and exhibits a large dispersion. Our results not only confirm that the remnants of GRBs are standard candles, but also provide the first evidence that the physical parameters of relativistic shocks are universal, with the favored values $\epsilon_{e}\sim 0.1$ and $\epsilon_{B}\sim 10^{-3}$.Comment: 11 pages including 3 color figure