Software Development Engineer - AI Workload Optimization
amdAustin (TX)
Software Development Engineer - AI Workload Optimization
Posted today
amdAustin (TX)
SENIORITY
Lead
About the role
Overview:
In this role you will define and drive the end-to-end software optimization strategy to achieve industry-leading performance for top-tier AI customers. You operate at the intersection of architecture, customer engagement, and software engineering to tune AMD’s software stack—from ROCm and compilers to high-level AI frameworks—for demanding workloads. You’ll shape roadmaps, mentor engineers, and partner with customers to deliver architectural wins. This is a strategic, impact-driven leadership position in a fast-paced, collaborative environment.
Responsibilities:
Set the technical vision and roadmap for workload optimization across the AI software stack Lead profiling, analysis, and tuning of large-scale models (LLMs, Diffusion, Multimodal, MoE) for AMD hardware Partner with top customers and hyperscalers to understand workload requirements and deliver tailored optimizations Collaborate across hardware, compiler, and framework teams to influence future silicon features Drive development of tools and frameworks for performance estimation, modeling, and reporting Act as a technical ambassador in industry forums and open-source communities; mentor future leaders
Key requirements:
Deep software development experience with high-level technical leadership (Fellow or equivalent) Deep expertise in AI frameworks (PyTorch, JAX, vLLM, SGLang) and ROCm stack Proven history of optimizing distributed inference and training at scale across multi-node/multi-GPU Mastery of performance profiling tools (TorchProfiler, ROCm Profiler, Nsight) and hardware-level performance modeling Strong understanding of modern model architectures (Transformer, Attention, KV Cache) and techniques (quantization, speculative decoding, FlashAttention) Ability to drive cross-functional initiatives in fast-paced, ambiguous environments Bachelor in Computer Science (or related) or equivalent experience Exceptional communication and ability to translate complex technical challenges into strategic roadmaps Cross-functional collaboration across hardware, software, and customer teams Mentorship and leadership of engineers and future technical leaders ROCm software stack PyTorch, JAX, vLLM, SGLang Profiling: TorchProfiler, ROCm Profiler, Nsight
Before you apply
Applying takes about a minute. These four things decide how fast it moves after that.
Your profile is current
It's what we read first. Occupations, seniority and locations matter more than a long history.
Two examples you can talk through
Not a portfolio — just two pieces of work where you can explain the decisions and what you'd change.
A number in mind
What you're on now and what would make you move. We negotiate better when we know both.
Your notice period
Employers plan around it, and it's the question that stalls offers most often.
Once you apply, someone reads it and calls you before anything reaches the employer — usually within two working days.