GPU Performance Architect

Job Title

GPU Performance Architect

Role Summary

Work with architecture, software, and design teams in RTG to model, analyze, and prototype GPU performance for data center and HPC products. Drive performance projections, validate architectural ideas through simulation and kernel prototyping, and provide actionable recommendations for next-generation hardware and software.

Experience Level

Senior. Years of experience not explicitly stated in the posting.

Responsibilities

The role focuses on performance modeling, workload-driven analysis, prototyping across the stack, and communicating results to architects and design teams.

• Develop high-level and detailed performance models for GPU subsystems, SoCs, and multi-GPU systems.
• Build analytical, simulation-based, and workload-driven models to predict application performance.
• Generate performance projections for future architectures and product configurations; establish KPIs for performance, latency, throughput, and efficiency.
• Prototype and validate ideas across software, firmware, and architectural models.
• Write and optimize GPU kernels and microbenchmarks for performance verification and competitive analysis.
• Collect, summarize, and present simulation and experiment results to architecture and design teams.
• Collaborate with architects to propose implementable hardware and software solutions.

Requirements

Must-have technical skills and constraints, plus preferred experience listed separately.

• Demonstrated experience with GPU architecture and performance analysis.
• Experience building analytical or cycle-accurate performance simulators or models.
• Strong programming skills in C/C++ and scripting (Python); ability to write and optimize GPU kernels (assembly, HIP, Triton, etc.).
• Familiarity with ML frameworks (TensorFlow, PyTorch) and common ML operators (GEMM, Attention) for workload analysis.
• Understanding of compute APIs such as CUDA and OpenCL.
• Proven ability to design experiments, collect data, and produce concise technical summaries for architects and design teams.
• Must be authorized to work in the United States; this role is not eligible for visa sponsorship.

Nice-to-have:

• Background in NoC/interconnect systems and multi-GPU scaling.
• Experience with compiler stacks (graph compilers, MLIR, LLVM) or building ML operator libraries.
• Experience with hardware modeling/design (RTL, SystemC) is a plus.

Education Requirements

Undergraduate degree required. The posting prefers a BS, MS, or PhD with emphasis in Electrical Engineering, Computer Architecture, Computer Engineering, or Computer Science. Bachelor’s or Master’s in computer engineering/Electrical Engineering explicitly cited; advanced degrees preferred but not strictly required.

About the Company

Company: Advanced Micro Devices

Headquarters: Sunnyvale, California, USA

Advanced Micro Devices, or AMD, is a global semiconductor company that designs and manufactures microprocessors, graphics processors, and related technologies for a variety of computing devices. Known for pushing the boundaries of innovation, AMD's mission is to deliver high-performance computing solutions for AI, data centers, gaming, and embedded applications. They foster a collaborative, inclusive culture focused on creativity and problem-solving, aiming to drive progress and excellence in technology.