Speed up deep learning workload performance on Intel CPUs and GPUs using Model Zoo’s optimized inference apps and the Intel Extension for TensorFlow, then analyze and debug the results.

This session focuses on the performance of your TensorFlow models—deep neural networks used for things like speech recognition, image processing, sentiment analysis, and even drones.

AI Software Solutions Engineer Louis Tsai will unpack the performance benefits of the Intel Extension for TensorFlow when used with Model Zoo for Intel Architecture.

Takeaways:

• An overview of Intel’s TensorFlow extension and its latest optimizations
• Analysis of performance improvements using TensorFlow Profiler
• How to benchmark your specific Intel platforms using Model Zoo workloads for TensorFlow