In this session we will cover various technologies from AMD has enabled unprecedented performance on High-Performance Computing (HPC) and Mega Data Center (MDC) applications. We will provide details of the second generation EPYCTM server CPU, performance analysis and profiling tools, and examples of workload optimization to maximize delivered performance.
Session Time: December 19, 3:30 pm
The 2nd generation AMD EPYCTM server CPU
We will kick off the session with an overview of AMD’s 2nd generation EPYCTM server CPU (codenamed “Rome”), featuring upto 64 high-performance Zen2 based cores, and go over highlights of the core micro-architecture and chiplet-based design.
Speakers
Anasua Bhowmik, Principal Manager of Technical Staff, Silicon Design Engg
Kanishka Lahiri, Principal Member of Technical Staff, Silicon Design Engg
Partner Talk
Data Center Trends and Performance Tuning For Large-Core Count Servers At Scale
Speakers: Sundar Dev and Rama Govindaraju, Google
AMD for High-Performance Computing and Machine Learning
We will introduce products from AMD for HPC and ML. We will give an overview of current and roadmap of products and how we plan to meet the HPC/ML needs. We will introduce both hardware and software products/tools for HPC/ML.
Speaker: Prakash Raghavendra, AMD Fellow, Software Dev Engg
AMD uProf: Know Your Performance Metrics
This session will provide an overview of AMD uProf, a performance, power and system analysis tool by AMD.This tool allows developers to better understand the runtime performance of their application and to identify ways to improve its performance. Session will include the basic concepts of profiling, touch upon major features of AMD uProf and how to use the tool on AMD platforms.
Speaker: Rajeeb Barman, Principal Member of Technical Staff, Software Dev Engg
