GSvit documentation

open source FDTD solver with GPU support

User Tools

Site Tools


start:tests

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision
Previous revision
start:tests [2015/07/17 22:32]
pklapetek
start:tests [2019/01/17 09:48] (current)
dnezval
Line 12: Line 12:
   * **./gsvit test 0** checks available GPUs only, printing their main characteristics.   * **./gsvit test 0** checks available GPUs only, printing their main characteristics.
   * **./gsvit test 1** makes single test on CPU and all GPUs (if there are any)   * **./gsvit test 1** makes single test on CPU and all GPUs (if there are any)
-  * **./gsvit test 2** tests key algorithms performance,​ running approximately ten different tests for 100x100x100 voxels, 100 steps, which takes some ten minutes. +  * **./gsvit test 2** tests key algorithms performance,​ running approximately ten different tests for 100x100x100 voxels, 100 steps, which takes about ten minutes. 
-  * **./gsvit test 3** tests key algorithms performance,​ running approximately ten different tests for 200x200x200 voxels, 300 steps, which can take tens of minutes, however +  * **./gsvit test 3** tests key algorithms performance,​ running approximately ten different tests for 200x200x200 voxels, 300 steps, which can take tens of minutes, however ​it is more precise regarding peak GSvit performance estimation.
-is more precise regarding peak GSvit performance estimation.+
   * **./gsvit test 12** compares GPU/CPU time scaling up to 200x200x200 voxels   * **./gsvit test 12** compares GPU/CPU time scaling up to 200x200x200 voxels
   * **./gsvit test 13** compares GPU/CPU time scaling up to 300x300x300 voxels   * **./gsvit test 13** compares GPU/CPU time scaling up to 300x300x300 voxels
Line 28: Line 27:
     * E - run Near-to Far Field transformation for a single point: 0 - no, 1 - yes     * E - run Near-to Far Field transformation for a single point: 0 - no, 1 - yes
     * F - source: 0 - point source, 1 - total/​scattered field, 2 - scattered field     * F - source: 0 - point source, 1 - total/​scattered field, 2 - scattered field
-The "​AAABCDEF"​ naming is used also to store the precomputed data that are compared with tests results. The precomputed +   
-data for comparison are typically single near field or far field point time dependence.+The "​AAABCDEF"​ naming is used also to store the precomputed data that are compared with tests results. ​ 
 +The precomputed data for comparison are typically single near field or far field point time dependence.
  
 There are two more tests designed for benchmarking HPC systems: There are two more tests designed for benchmarking HPC systems:
start/tests.1437165139.txt.gz · Last modified: 2018/01/24 08:14 (external edit)