Browsing by Author "Yang, Bo"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
- ItemOpen AccessA GMRES Solver with ILU(k) Preconditioner for Large-Scale Sparse Linear Systems on Multiple GPUs(2015-09-28) Yang, Bo; Chen, Zhangxing (John)Most time of reservoir simulation is spent on the solution of large-scale sparse linear systems. The Krylov subspace solvers and the ILU preconditioners are the most commonly used methods for solving such systems. Based on excellent parallel computing performance, GPUs have been a promising hardware architecture. The work of developing preconditioned Krylov solvers on GPUs is necessary and challengeable. We devote our efforts into the development of the GMRES and the ILU(k) preconditioner on a multiple-GPU architecture and achieve favorable speedup effects. Our GPU computation includes the algorithms such as SPMV, nested RAS, decoupled ILU(k) and parallel triangular solver, etc. The numerical experiments prove that our preconditioned GMRES algorithm is feasible and works well on a multiple-GPU workstation.
- ItemOpen AccessPreconditioned Iterative Solvers on GPU and an In-Situ Combustion Simulator(2020-01) Yang, Bo; Chen, Zhangxing; Liao, Wenyuan; Lu, Qingye; Azaiez, Jalel; Sun, Pengtao; Ware, Antony FrankThis thesis consists of two parts: preconditioned iterative solvers on GPU and an in-situ combustion simulator. The purpose of the first research is to develop a new parallel solution platform based on GPU features. An application of HPC (high-performance computing) technology to reservoir simulation has become an inevitable trend. As a platform for HPC, GPU can provide an effective solution for personal computers and workstations. In this research, not only a series of special CPR (constrained pressure residual) preconditioned solvers are developed for black oil models, but also a variety of other preconditioned solvers are completed as contrast solvers. The numerical experiments verify a significant improvement in the parallel performance of the solvers on GPU. They also provide an overall comparison among the combinations of different GPUs, solvers, and preconditioners. The results demonstrate that the CPR developed has excellent advantages in both parallelism and convergence for the solution of a benchmark reservoir model. The purpose of the second research is to develop a new comprehensive ISC (in-situ combustion) simulator with the PER (pseudo equilibrium ratio) method and to compare functions with those of a benchmark simulator with the VS (variable substitution) method. ISC is considered a promising recovery method because of its low cost and less environmental impact. However, an ISC simulator is regarded as one of the most complex simulators to develop. The PER method can reduce the complexity of simulator development because it lowers the influence of the phase disappearance and appearance on the mathematical system of reservoir simulation. The ISC simulator in this study is developed with comprehensive typical functions of the ISC process. For the verification of the equivalence in numerical results between the PER method and the VS method, the numerical experiments are carried out in an omnidirectional range. Because the results show a very close match, the research provides reliable experimental support for popularizing the use of the PER method to develop an in-situ combustion simulator.