Principal AI Performance Engineer - LLM Inference (SGLang)

AMD; Helsinki; On-site; EUR 90,000 - 150,000; Full time

Job summary

AMD is looking for a performance-obsessed engineer to drive AI inference performance to the absolute limit on AMD GPUs, with SGLang as the primary serving framework. You will lead a small, highly technical team and work end-to-end across the stack: profiling, diagnosing, and optimizing leading models running on SGLang across customer-relevant serving configurations (e.g.

agentic coding, long-context, high-throughput serving).

Qualifications

• 7+ years of software development experience in GPU computing or AI systems.
• Hands-on experience with SGLang internals; familiarity with vLLM or TensorRT-LLM is a plus.
• Strong workload profiling and bottleneck diagnosis skills.
• Understanding of GPU kernel performance characteristics and model architectures.

Responsibilities

• Drive performance optimization end-to-end on SGLang across models and configurations.
• Profile, diagnose, and resolve hardest cross-stack bottlenecks in SGLang deployments.
• Diagnose kernel-level issues and translate findings into optimizations.
• Lead customer-facing technical engagements and present measurable uplifts.
• Integrate and optimize custom kernels within SGLang (HIP, CUDA, Triton, CK, PyDSL, ASM, AITER).
• Optimize multi-node distributed inference and scale-out performance.
• Develop and refine shared performance optimization methodologies.
• Leverage AI agents to accelerate daily work and define best practices.
• Contribute upstream to SGLang and open-source frameworks such as vLLM and PyTorch.

Skills

Python

C++

Linux systems

Customer-facing leadership

GPU computing

Profiling

English communication

Education

Master's degree in Computer Science/Engineering

PhD preferred

Tools

HIP

CUDA

Triton

CK

Gluon

PyDSL

ASM

AITER

RCCL/NCCL

RDMA

Job description

AMD is looking for a performance-obsessed engineer to drive AI inference performance to the absolute limit on AMD GPUs, with SGLang as the primary serving framework. You will lead a small, highly technical team and work end-to-end across the stack: profiling, diagnosing, and optimizing leading models running on SGLang across customer-relevant serving configurations (e.g. agentic coding, long-context, high-throughput serving).

KEY RESPONSIBILITIES

• Drive performance optimization end-to-end on SGLang across leading models and customer-relevant serving configurations, closing competitive gaps through kernel and systems-level optimizations
• Profile, diagnose, and resolve the hardest cross-stack performance bottlenecks in SGLang deployments, from GPU kernels and operator dispatch to the SGLang scheduler, RadixAttention/prefix caching, and multi-node communication
• Diagnose kernel-level performance issues using profiling tools: identify occupancy limitations, L2 cache thrashing, register pressure, memory coalescing issues, etc, and translate findings into actionable optimizations
• Lead customer-facing technical engagements: present findings, recommend optimizations, and deliver measurable performance uplifts on SGLang
• Integrate and optimize custom kernels (Triton, Gluon, CK, PyDSL, ASM, AITER) within SGLang, understanding dispatch paths, shape extraction, and backend selection
• Optimize multi-node distributed inference on SGLang: communication-compute overlap, parallelism strategies (TP/EP/DP), and scale-out performance
• Develop and refine shared performance optimization methodology that raises the bar across the broader team
• Leverage AI agents to accelerate daily work and define best practices for AI-assisted performance engineering
• Upstream optimizations into SGLang and adjacent open-source frameworks such as vLLM and PyTorch

PREFERRED EXPERIENCE

• 7+ years of software development experience in GPU computing, AI systems, or high-performance computing
• Deep hands-on experience with SGLang internals; familiarity with vLLM, TensorRT-LLM, or similar is a plus
• Strong background in end-to-end workload profiling and bottleneck diagnosis
• Understanding of GPU kernel performance characteristics: occupancy, register and LDS pressure, memory coalescing, cache utilization, wavefront scheduling, and instruction-level bottlenecks
• Understanding of model architectures (transformers, MoE, diffusion), inference paradigms (speculative decoding, prefill-decode disaggregation, continuous batching)
• Experience with custom kernel development or integration (HIP, CUDA, Triton, CK, or similar)
• Understanding of multi-GPU and multi-node distributed systems: RCCL/NCCL, RDMA
• Strong proficiency in Python and C++
• Customer-facing technical leadership experience
• Fluent in AI-assisted development
• Strong Linux systems knowledge
• Excellent written and verbal English communication skills

ACADEMIC CREDENTIALS

Master's, or PhD in Computer Science, Computer Engineering, Electrical Engineering, or equivalent. 7+ years of software development experience in GPU computing, AI systems, or high-performance computing, Deep hands-on experience with SGLang internals, Strong background in end-to-end workload profiling and bottleneck diagnosis, Understanding of GPU kernel performance characteristics (occupancy, register/LDS pressure, memory coalescing, cache utilization), Understanding of model architectures (transformers, MoE, diffusion) and inference paradigms (speculative decoding, continuous batching), Experience with custom kernel development or integration (HIP, CUDA, Triton, CK, or similar), Understanding of multi-GPU and multi-node distributed systems (RCCL/NCCL, RDMA), Strong proficiency in Python and C++, Customer-facing technical leadership experience, Strong Linux systems knowledge, Excellent written and verbal English communication skills, Master's or PhD in Computer Science, Computer Engineering, Electrical Engineering, or equivalent