Easily Accelerate AI With FPGAs (Internet of Things Applications USA 2019)

Dr Ted Way, Senior Program Manager, AI and Advanced Architectures
United States


USA 2019 _ Microsoft _ Presentation*
USA 2019 _ Microsoft _ Presentation*
USA 2019 _ Microsoft _ Presentation*

If you already have access, please [Login]


Presentation Summary

Computer vision is powered by deep neural networks, which are computationally expensive. In this talk we will show how data scientists can use familiar tools such as Python and TensorFlow to train a model, and then accelerate it on an FPGA. The models are containerized and can be run on FPGAs in the Azure cloud or on Data Box Edge devices on-prem. The entire workflow is integrated with Azure Machine Learning, a comprehensive data science platform.

Speaker Biography (Ted Way)

Ted Way is a senior program manager on the AI and Advanced Architectures team at Microsoft, where he works on hardware acceleration of AI in the cloud and edge. He's passionate about telling the story of how AI will empower people and organizations to achieve more. He holds BS degrees in electrical engineering and computer engineering, MS degrees in electrical engineering and biomedical engineering, and a PhD in biomedical engineering from the University of Michigan—Ann Arbor. His PhD dissertation was on "spell check for radiologists," a computer-aided diagnosis (CAD) system that uses image processing and machine learning to predict lung cancer malignancy on chest CT scans.

Company Profile (Microsoft)

Microsoft is simplifying IoT so every business can digitally transform through IoT solutions that are more accessible and easier to implement. Microsoft has the most comprehensive IoT portfolio with a wide range of IoT offerings to meet organizations where they are on their IoT journey, including everything businesses need to get started — ranging from operating systems for their devices, cloud services to control them, advanced analytics to gain insights, and business applications to enable intelligent action.
View Microsoft Timeline