Arch Linux NVIDIA GPU CLOUD 18.04-py3 tensorflow googlen TITAN V1枚 での学習能力を確認してみた Images/sec 1349.7

TensorFlow: 1.7.0
This script: nvcnn.py v1.6
Cmd line args:
–model=googlenet
–batch_size=256
–num_gpus=1
–fp16
Num images: Synthetic
Model: googlenet
Batch size: 256 global
256.0 per device
Devices: [‘/gpu:0’]
Data format: NCHW
Data type: fp16
Have NCCL: True
Using NCCL: True
Using XLA: False
Building training graph
Creating session
●Images/sec: 1349.7●
Arch Linux TITAN V NVIDIA GPU CLOUD 18.04-py3 tensorflow googlenet 学習能力確認 Images sec 1349.7
GPU nvidia-smi
Arch Linux TITAN V NVIDIA GPU CLOUD 18.04-py3 tensorflow googlenet 学習能力確認 Images sec 1349.7 nvidia-smi