Intel AI
Keynote Stephan Gillich
Development steps Deployment Acceleration AI Software AI Community Optimized Frameworks/Toolkits DL Framework
Optimized deep learning frameworks
Neural network -- how to decide
Inception v3
VGG16
Mobilenet
Intro
Current hardware: Cascade Lake
DevCloud
Hands-on
Individually assigned IP-address for each participant for hands-on ...
get Intel-optimized TensorFlow
ipynb2: 299,299,64
Batch: bigger = faster but potentially less accurate (Paper)
David Brayford (LRZ) Charliecloud @ LRZ
Guest talk:Intel ML frameworks
Charliecloud
Tutorial / Demo
Tools available convert Docker to
Scikit Learn
DAAL4py k-means MKL for numpy Kernel PCA
DevCloud
software.intel.com/edge