SoC System Performance Architect AI/ML

Role Overview

The SoC System Performance Architect specializes in the design and optimization of high-performance System on Chips (SoCs) specifically for cloud computing and machine learning applications. The individual will be responsible for developing innovative solutions that enhance performance and optimize power efficiency in next generation data center systems.

Position Summary

As part of AMD's Data Center GPU organization, the role involves collaborating closely with a team of engineers dedicated to advancing AI-powered computing solutions. The key focus is on architecting, analyzing, and optimizing SoCs to solve complex performance challenges and contribute to cutting-edge technology development.

Experience Level

This position requires a seasoned professional with demonstrated knowledge in system performance architecture within the computing industry, ideally having prior experience related to GPUs or similar technologies.

Key Duties

The main responsibilities of this role include:

• Acting as a Performance Architect in the electronics design engineering division.
• Undertaking microarchitecture exploration, performance modeling, and analysis of next-gen GPU systems.
• Understanding design architecture in-depth and proposing enhancements for performance.
• Conducting micro-benchmarking, workload characterization, competitive analysis, and identifying performance bottlenecks.
• Creating and employing tools and methodologies for the performance analysis of various workloads.
• Communicating and executing solutions to address processor and system performance issues.

Candidate Requirements

Successful candidates will demonstrate the following expertise:

• Strong background in GPU, CPU, or computer system microarchitecture.
• Experience in developing and implementing computer system performance models.
• Proficient in performing computer workload analysis.
• Strong programming skills, particularly in C++ and Python (or similar languages).

Education Requirements

A PhD in Computer Engineering, Electrical Engineering, Computer Science, or a related field is required. Alternatively, a Master's degree in one of these fields combined with significant experience as a Computer Architect or Design Engineer will be considered sufficient.