Fellow, AI Workload Optimization
Core
Define and drive end-to-end software optimization strategy for AMD's AI software stack to achieve industry-leading performance for top-tier customers.
Role type
Senior IC Fellow (AI Software Optimization)
Builds
AMD's software stack (ROCm, compilers, high-level AI frameworks) tuned for maximum performance on demanding AI workloads.
Domain
Semiconductor / AI Infrastructure
Deliverable
production ML models
Required skills
AI framework expertise (PyTorch, JAX, vLLM, SGLang), ROCm stack mastery, distributed inference/training optimization, performance profiling (TorchProfiler, ROCm Profiler, Nsight), modern model architecture knowledge (Transformer, Attention, KV Cache), optimization techniques (quantization, speculative decoding, FlashAttention), hardware-software co-design, strategic roadmap creation, customer engagement with hyperscalers, open-source community leadership.
Preferred skills
PhD or Master's in CS/EE, deep research/applied experience in AI/ML (deep learning, LLMs, computer vision), experience in high-level technical leadership (Fellow or equivalent).
Technologies
ROCm, PyTorch, JAX, vLLM, SGLang, TorchProfiler, Nsight, FlashAttention
Responsibilities
Set technical vision and roadmap for workload optimization; lead profiling and tuning of large-scale models (LLMs, Diffusion, Multimodal, MoE); partner with customers and hyperscalers to deliver tailored optimizations; collaborate with hardware/compiler teams to influence silicon features; drive development of performance estimation tools; mentor technical leaders and engineers.
Seniority
Fellow, strategic technical leadership
