This release focuses on accelerating deep learning inference workloads on Intel® Xeon® (CPU processor) and has the following key features:
- Out-of-box installation experience for TensorFlow*, MXNet*, and ONNX.
- Validated optimizations for 17 workloads each on both TensorFlow and MXNet, as well as 14 for ONNX.
- Support for Ubuntu 16.04 (TensorFlow, MXNet and ONNX).
- Support for OSX 10.13.x (buildable for TensorFlow and MXNet).
This 0.15 release includes optimizations built for popular workloads already widely deployed in production environments. These workloads cover the following categories:
image recognition & segmentation
speech generation & recognition
Generative Adversarial Networks (GAN)
In our tests, the optimized workloads can perform up to 45X faster than native frameworks, and we expect performance gains for other workloads due to our powerful Pattern matcher feature.
See also our recent API changes