Hardware/Software co-design for AI accelerators

23 Aprile 2024
Share on facebook Share on twitter
200
VIEW

Alessandro Palla (Intel)
23 aprile
dalle 10:30 alle 13:30
aula SI.3 del Polo B

Link Teams

 

Abstract
AI applications are increasingly shifting from the cloud to users' personal PCs. The integration of edge AI accelerators necessitates finely tuned hardware-software co-design to fulfill customer demands for high performance while maintaining an extremely low-power design. This talk will provide insights into the workings of neural networks on AI accelerators, highlighting essential hardware-software trade-offs and compiler optimizations necessary to meet stringent performance/watt requirements.

Bio
Alessandro Palla is a staff deep learning engineer at Intel. He graduated in electronics engineering, and got the related PhD, in 2014 and 2018 respectively at the University of Pisa. He is working since 2017 in Intel Corporation, designing next generation Neural Processing Units (NPU) AI accelerator on Intel Client CPU. His domain of expertise is hardware/software codesign and compiler optimization techniques for AI accelerators.