Administrator Guide

10 RAPIDS Scaling on Dell EMC PowerEdge Servers
2 System Configuration
Test System Hardware:
Servers
• C4140-M
o 4xV100-SXM2-16GB
• R940xa
o 4xV100-PCle-32GB
o 4xV100-PCle-16GB
• Network connection over InfiniBand
• R740xd server hosting the NFS with the dataset for remote data
Test System Software:
• Ubuntu 18.04
Docker CE v19.03+ for Linux distribution
• RAPIDS Version: 0.10.0 - Docker Install
Nvidia Driver: 418.67+
CUDA: 10.1
Note: The systems were tested using RAPIDS via docker NVIDIA GPU Cloud (NGC); the original
notebook requires Google DataProc to access the hosted NYC-Taxi Dataset; to bypass this dependency
we downloaded the dataset locally and update the notebook to read if from a specific path.
Test Configuration:
The tests were conducted using the below variations;
• Using dataset from year 2014 until year 2016, exploring the maximum dataset capacity to be handled
by each server.
• RMM enable and disable
• Single Node [3-4] and Multi Node [5-8]
• Local data and remote data on NFS [9]
• To ensure reproducibility, each server was tested 3 times and calculated the average as the result.