This is the code used for a few of the blog posts on: https://lambdalabs.com/blog including:

• 2080 Ti Deep Learning Benchmarks: https://lambdalabs.com/blog/2080-ti-deep-learning-benchmarks/

Environment:

• OS: Ubuntu 18.04
• TensorFlow version: 1.11.0-rc1
• CUDA Version 10.0
• CUDNN Version 7.3.0

Step One: Clone benchmark repo

cd ~
git clone https://github.com/lambdal/benchmarks.git
git clone https://github.com/lambdal/lambda-tensorflow-benchmark.git

Step Two: Run benchmark

Note: This script assumes the lambdal/benchmarks repo is inside of your home directory. If you've cloned it elsewhere, you need to change the SCRIPT_DIR variable in lambda-tensorflow-benchmark/script_benchmark.sh accordinly.

cd lambda-tensorflow-benchmark
./script_benchmark.sh

Step Three: Report results

./script_report.sh