A5039273288- Distributed and Parallel Computing Systems
- Parallel Computing and Optimization Techniques
- Particle Accelerators and Free-Electron Lasers
- Particle accelerators and beam dynamics
- scientometrics and bibliometrics research
- Mobile Agent-Based Network Management
- Advanced Computational Techniques and Applications
- Digital Transformation in Industry
- Modular Robots and Swarm Intelligence
- Power Systems and Technologies
- Advanced Data Storage Technologies
- Robotics and Automated Systems
- Superconducting Materials and Applications
- Magnetic confinement fusion research
- Cloud Computing and Resource Management
- Matrix Theory and Algorithms
Chinese Academy of Sciences
2010-2024
Institute of High Energy Physics
2024
University of Chinese Academy of Sciences
2024
China Spallation Neutron Source
2024
Karlsruhe Institute of Technology
2022
University of Surrey
2022
Shenyang Institute of Automation
2022
Arizona State University
2022
John Wiley & Sons (United States)
2022
Hudson Institute
2022
Sparse Matrix-Vector multiplication (SpMV) is an important computational kernel in scientific applications. Its performance highly depends on the nonzero distribution of sparse matrices. In this paper, we propose a new storage format for diagonal matrices, defined as Compressed Row Segment with Diagonal-pattern (CRSD). CRSD, design patterns to represent distribution. As Graphics Processing Units (GPUs) have tremendous computation power and OpenCL makes them more suitable computing, implement...
In large-scale cluster systems, interconnecting thousands of computing nodes increase the complexity network topology. Nevertheless, few existing computational models consider impact hierarchical communication latencies and bandwidths caused by complexity. this paper we propose a new parallel model called LogGPH with parameter H incorporated into LogGP to describe hierarchy. Through predicting analyzing point-to-point collective MPI_Allgather on two 100-Terascale supercomputers, Dawning...
In this paper, we present our early performance evaluation results with the NPB benchmark and two scientific computing applications program, i.e., a HFFT package developed by lab CFDO application software, on 100 Teraflops-scale Dawning 5000A DeepComp 7000. We compared of 7000, their corresponding predecessor, 4000A 6800, which demonstrating improvements across variety problems. From results, can find that keep its scalability up to 16384 cores while 7000 number becomes 4096. also scales...
Abstract The design betatron tune of the Rapid Cycling Synchrotron (RCS) China Spallation Neutron Source (CSNS) is (4.86, 4.80), which allows for incoherent shifts to avoid serious systematic resonances. When operational bare was set at value, beam instability in horizontal plane and loss induced by half-integer resonance vertical under space charge detuning were observed. Simulations experiments have shown that charge-induced reduces as tunes move up away from lines. However, experimental...