GPU Performance Architect
Core
Analyze GPU bottlenecks, simulate performance gains, and propose architectural improvements for graphics and compute workloads.
Role type
Senior IC GPU Performance Architect
Builds
Next-gen GPU hardware solutions and world-class graphics processors
Domain
Semiconductor hardware design, GPU architecture, AI/ML compute
Deliverable
production ML models
Required skills
GPU architecture analysis, RTL and high-level simulation, C/C++ programming, Graphics/Compute APIs (DirectX/Vulkan/CUDA/HIP), Kernel development for LLMs, PPA optimization, Graphics profiling tools
Preferred skills
AI/ML algorithms, PyTorch/Triton/CUDA/HIP kernel development, Graphics/Compute compilers, GFX profiling tools (PIX, RenderDoc)
Technologies
PyTorch, Triton, CUDA, HIP, DirectX, Vulkan, RTL simulators, AMD tools
Responsibilities
Workload and competitive analysis for game/AI/ML applications, Identify and solve complex technical problems in graphics cores and SoCs, Analyze graphics/compute paradigms and algorithms, Implement and run simulation models to estimate performance gains, Propose innovative solutions for Power/Performance/Area (PPA) improvements, Drive initiatives to improve analysis tools
Seniority
Senior, hands-on IC
