02 Feb, 2016 Presentation
AccDNN is a seamless bridge between deep learning and FPGAs for data scientists. This bridge does not require RTL programming. AccDNN uses the basic neural network definition file and trained model weights (caffe model) to generate an FPGA implementation, and then automatically creates the recognition service in a cloud. This talk will describe the framework and process to convert the network definition file to Verilog for the FPGA. It will also show use cases and examples, as proven on SuperVessel today.
Yonghua Lin is the founder and leader of SuperVessel OpenPOWER cloud. Meanwhile, she is the Senior Technical Staff Member and Senior Manager of Cloud Infrastructure group in IBM Research. She has worked on system architecture research in IBM for 12 years. Her work covered all kinds of IBM multicore processors in the pass 10 years, including IBM network processor, IBM Cell processor, PRISM, IBM POWER 6, and POWER 7, etc. She was the initiator of mobile infrastructure on cloud from 2007 which has become the Network Function Virtualization today. She led IBM team built up the FIRST optimized cloud for 4G mobile infrastructure, and successfully demonstrated in ITU, Mobile World Congress, etc. She was the founder of SuperVessel cloud to support OpenPOWER research and development in industry. She herself has more than 50 patents granted worldwide and publications in top conferences and journals.
Jun Song Wang is the Research Staff Member in IBM Research, China. He has worked on wireless communications, signal processing and machine learning research in IBM for more than 5 years. He was one of the major contributors of the long distance wireless communication platform, and the initiator of the industrial atmospheric particle sensor for Green Horizon. Currently, he is working on the heterogeneous computing research in cloud, especially for the FPGA acceleration of deep learning. He has more than 20 patents granted worldwide and publications in top conferences and journals.