Develop machine learning applications on Cortex-A at Arm TechCon 2018

Now is the time to register for Arm TechCon 2018, happening October 16-18 at the San Jose Convention Center. The agenda for the conference is now available and includes over 70 hours of sessions with seven technical tracks.

Accelerating & Optimizing Machine Learning on Arm Cortex-A

One of the highlights is a hands-on workshop with Arm Cortex-A CPU and Mali Graphics technology titled Accelerating & Optimizing Machine Learning on Arm Cortex-A on Wednesday, October 17, at 3:30 PM.

The workshop is a two-hour, hands-on session introducing the 96Boards HiKey 960 development board. It covers how to create images and install the Ubuntu Linux distribution for Arm as well as how to build and run machine learning examples based on the Arm Compute Library and the Arm NN software developer kit. The first 40 attendees will receive a free board to take home and continue learning.

Machine learning (ML) application development is typically done using frameworks like TensorFlow, Caffe, Caffe2, or PyTorch. Python is often used to create and train deep learning models. For mobile and embedded applications, the created models need to be transferred to client and embedded hardware and executed on Arm CPUs and GPUs. The workshop explores how Arm software can help machine learning applications perform better on Arm hardware.

Arm Streamline is a performance analysis tool that can help software engineers optimize ML on Arm. During the workshop the performance of ML applications running on the HiKey 960 will be explored. This is a great way to develop and prototype on the latest Arm technology and analyze application performance.

Who should register for the workshop?

Perfect for engineers who want to learn about:

• Embedded Linux on a high-performance Arm development platform
• Available software from Arm to build ML applications
• Performance analysis of Linux applications running on Cortex-A CPUs and Mali GPUs and how to find performance bottlenecks
• Trade-offs between using CPU and GPU hardware for ML

Attendees receive:

• A free HiKey 960 development board
• A free 60-day license for Arm DS-5 Ultimate Edition, including Streamline
• A memory stick with software tools and workshop materials

How to register for the workshop at Arm TechCon

If you enjoy embedded Linux on Arm and are curious about how to create ML applications put the workshop on your TechCon calendar. Make sure to get there early to secure your spot and free HiKey 960 board. Attendees should be familiar with embedded Linux and have some experience with the Linux command line, scripting, and C/C++ development. Save the date and register today by clicking on the link below.

Register for TechCon