FPGA Acceleration of AI Computation
Published:
This is the presentation slides of course CSC6031: Advanced Computer Architecture instructed by Prof. Wei-Chung Hsu（徐慰中教授）. The presentation is a survey of how to use FPGA, a domain-specific architecture (DSA) to accelerate AI computation, and mainly conducted a case-study of Microsoft’s Catapult and Brainwave projects, explaining how they designed their FPGA chips to accelerate computation in Azure.