At the last Linaro Connect, Linaro hosted a workshop on Arm NN.
The workshop gave an overview of Arm NN, providing attendees with an introduction to the Arm NN Graph Builder, optimizing and running the network, backends and workloads, custom backends for third party NN processing solutions - be it a network processor unit, a deep learning accelerator or a heterogeneous computing architecture.
The workshop was followed by training throughout the week. This was an opportunity for developers to provide feedback directly to Arm - feedback which has been taken on board and in most cases incorporated in new features.
The next Linaro Connect will take place at Paradise Point in San Diego on 23-27 September 2019. Here we plan to host a follow up session by Arm to introduce new features and take on more feedback for future improvements.
Following the session, we will have a developer room set up where all attendees can collaborate on AI.
Bring you NN enabled board and develop your Arm NN plug-in with the experts!
We are truly excited to share that we will have special guests from Microsoft and Amazon AWS: Weixing Zhang will talk about Microsoft ONNX AI runtime and Animesh Jain will talk about the end-to-end graph compiler stack provided by AWS Sagemaker Neo.
Last but not least, Marc Charlebois, Markus Levy and Tom Curran, senior experts from Qualcomm, NXP and Avnet will provide technical talks about running optimised AI inference on Arm by leveraging the multiple NN processing units in their SoCs.
For more information about Linaro Connect or to view the complete schedule, go to https://connect.linaro.org/schedule/.
See below schedule for all Arm NN related sessions: