OpenFOAM并行测试
-
楼上的说法我可以提供一点佐证。处理器是至强E5-2679v4,40核心80线程,32G内存。网格数记不清了,大概160到180万多面体-六面体混合网格。VOF,k-omega SST, Sliding Mesh。纵轴是时间,单位是秒,横轴是线程数,图是随手画的,领会精神就好:
-
@东岳 学校的cluster,志强 6142。用的slurm,但是我不会直接提交Allrun,所以就只测试了6核(在本地运行太多怕被网管揍)
snappyHexMesh, 67.79s
simpleFoam, 149s
不过说明上说这个机器不适合MPI运行,最开始的时候出了个warning[[51384,1],0]: A high-performance Open MPI point-to-point messaging module was unable to find any relevant network interfaces: Module: OpenFabrics (openib) Host: dragon2-ctrl0 Another transport will be used instead, although this may result in lower performance. NOTE: You can disable this warning by setting the MCA parameter btl_base_warn_component_unused to 0.
-
@东岳 1812
-
@东岳 另外一个cluster上更搞笑,处理器是 志强 5118@2.3G, OF4.1,也是本地运行,均为6核,不过太慢估计是有别的原因
snappyHexMesh 96.72s
simpleFoam ExecutionTime 250.07s, ClcokTime 661s。感觉前者比较可信,因为这是个登录和编译用的节点,估计用的人比较多,后台线程比较多了。
也是有个warning-------------------------------------------------------------------------- WARNING: No preset parameters were found for the device that Open MPI detected: Local host: lm3-m001 Device name: i40iw1 Device vendor ID: 0x8086 Device vendor part ID: 14290 Default device parameters will be used, which may result in lower performance. You can edit any of the files specified by the btl_openib_device_param_files MCA parameter to set values for your device. NOTE: You can turn off this warning by setting the MCA parameter btl_openib_warn_no_device_params_found to 0.
-
-
@东岳 我找个时间研究下怎么提交Allrun的script。。。
-
今天终于想起把鸽了这么久的事情干了一下了,今天测试的是5118 2.3G. OpenFOAM 4.1
因为是服务器,所以先运行了surfaceFeatureExtract,blockMesh, decomposePar 以后,再用slurm 分别提交snappyHexMesh 和simpleFoam,运行完snappyHexMesh 以后把Allrun里面的这两行也运行一下再提交patchSummary, potentialFoam 和simpleFoam。不过是log文件里面是分别计时的,所以应该没影响。ls -d processor* | xargs -I {} rm -rf ./{}/0 ls -d processor* | xargs -I {} cp -r 0.orig ./{}/0
snappyHexMesh
6C: 99.52 s
12C: 71.93 s
24C: 59.36s
simpleFoam
6C: 237 s
12C: 158 s
24C simpleFoam 时间是86s, 基本符合预期 -
再做了一个笔记本的测试,CPU i7-9750H + 16G, 性能模式, win10 下 WSL Ubuntu 18.04,OpenFOAM 7,6核,
第一遍 snappyHexMesh 74.98s, simpleFoam 251s
第二遍 snappyHexMesh 69.01s, simpleFoam 256s -
@hurricane007 志强金牌5118六核计算237s?