Advanced AI Workloads - Performance and Scalability Engineer
Core
Drive AI inference performance to the absolute limit on AMD GPUs by profiling, diagnosing, and optimizing leading models across the stack for customer-relevant serving configurations.
Role type
Senior IC performance and scalability engineer (AI workloads)
Builds
High-performance AI inference serving on AMD GPUs
Domain
AI systems, GPU computing, high-performance computing
Deliverable
production ML models
Required skills
GPU kernel optimization, distributed inference, profiling and bottleneck diagnosis, custom kernel integration, multi-node communication optimization, Python, C++, Linux systems, AI-assisted development
Preferred skills
AI serving frameworks (vLLM, SGLang, TensorRT-LLM), model architecture knowledge (transformers, MoE), system and rack-level design, customer-facing technical leadership
Technologies
HIP, CUDA, Triton, CK, PyDSL, ASM, AITER, vLLM, SGLang, PyTorch, NCCL, RDMA
Responsibilities
Drive end-to-end performance optimization across the stack; Profile, diagnose, and resolve cross-stack performance bottlenecks; Integrate and optimize custom kernels within serving frameworks; Optimize multi-node distributed inference; Lead customer-facing technical engagements; Upstream optimizations into open-source frameworks
Seniority
Senior, hands-on IC with team leadership
