/PAC2019

PAC2019 Optimization Group Silver Award

Primary LanguageFortran

PAC2019优化组决赛 手算队201

北京航空航天大学手算队 优化组银奖

优化内容:

  1. 编译选项
  2. 算法
  3. 缓存优化
  4. 精度优化

This readme file contains compilation and run scripts for the sample codes. Bin 10/22/2015 pcg_sfc_quiz: compile: compile_pcg.sh run : run_pcg.sh

Notices:

  1. Residual threshold is adjusted so that double-precision floating-point might be necessary for convergence.
  2. MPI rank number is adjusted to 24 so that no core is oversubscribed on HSW/ BDW/SKL/CLX processor.
  3. The result should be reproducible. Please submit all changes, including codes, building and running scripts.
  4. Can't modified the file: pcg.90 , solver_pcg_mod.f90 , time_management.f90, pop_in
  5. Algorithm optimization is strongly encouraged. Some references on Communication- avoiding Krylov subspace methods are given below, as well as the P-CSI (Chebyshev Iteration) paper. Some infrastructural functions (hints) are already provided, such as create_boundary_capcg and update_matpow_halo for thicker boundary update in CA-PCG solver, global_sum_1d_dbl for one-dimensional double array, etc.

References: [1] M. Hoemmen, Communication-avoiding Krylov subspace methods, PhD thesis, University of California, Berkeley, 2010. [2] Erin Carson, Communication-Avoiding Krylov Subspace Methods in Theory and Practice, PhD thesis, University of California, Berkeley, 2015. [3] Erin Carson and James Demmel, A Residual Replacement Strategy for Improving the Maximum Attainable Accuracy of s-step Krylov Subspace Methods. SIAM J. Matrix Anal. Appl., 35(1), 2014, pp. 22-43. [4] J. Demmel, M. Hoemmen, M. Mohiyuddin, and K. Yelick, Avoiding communication in computing Krylov subspaces, Tech. Report UCB/EECS-2007-123, EECS Dept., U.C. Berkeley, Oct 2007. [5] Yong Hu, Xiaomeng Huang, etal. Improving the Scalability of the Ocean Barotropic Solver in the Community Earth System Model. SC 2015.