Ceci est une ancienne révision du document !


Performance studies

The objective of this study is to analyze xgboost's performance on the cluster.

Prerequisites

  • Python Ananoda 3.7
  • Xgboost (Installed with conda)
  • Mesoshared (CPU Xeon Platinum 8160, 2.1 GHZ) With 92 cores

Covert the notebook to python script:

$ ipython nbconvert --to python 3.XGBoost
Xgboost is multithreaed. We control the number of threads by OMP_NUm_THREADS env variable

For example to for 8 cores:

$ OMP_NUM_TREADS=8 time ipython 3.XGBoost.py
# Threads Time(m)
111m32s
82m03s
161m18s
241m8s
321m5s
401m5s
481m5s
641m54s
722m13s
92
Conclusion : Uper limit scalability is about 24 cores

FIXME TODO

  • perf/xgboost.1562256303.txt.gz
  • Dernière modification : 2019/07/04 16:05
  • de admin