Introduction to GPU architecture and IREE's GPU CodeGen Pipeline | Online Only

Abstract:
This talk provides an introduction to GPU architecture. It further discusses foundational principles for designing intermediate representations (IR) using MLIR, specifically targeting GPUs. The session aims to build an intuitive understanding of code generation workflows and architectural considerations in modern GPU systems

Speaker Bio:
Prashant Kumar is a Member of Technical Staff at AMD working on IREE. Prior to AMD he was working as a Machine learning Compiler Engineer at Qualcomm. He holds a Masters degree in Computer Science from IIT Delhi. He is an active contributor to the IREE and FPL library

Github: https://github.com/pashu123
LinkedIn: https://www.linkedin.com/in/prashantkumar25/