LLM Inference Performance Engineer

Intel; Santa Clara (CA); Hybrid; USD 171,000 - 315,000; Full time

Job summary

Intel is seeking an AI Infrastructure Engineer in Santa Clara to push LLM inference on Intel GPUs, optimizing the end-to-end stack from profiling to kernel development. You will upstream improvements to open-source platforms like vLLM, SGLang, and PyTorch and collaborate with architecture teams to shape future GPU roadmaps.

The role requires strong C++/Python skills and 4+ years in GPU computing or HPC, with a passion for performance and scale-out inference across multi-node setups.

Qualifications

• Bachelor's degree in CS/SE/AI with 4+ years experience or higher degrees with corresponding years.
• 3+ years in GPU computing, AI systems or HPC.
• Proficiency in C++ and Python; able to read/modify complex systems code.

Responsibilities

• Own end-to-end optimization for running state-of-the-art LLMs on Intel GPUs.
• Profile, diagnose, and resolve cross-stack performance bottlenecks.
• Design and optimize custom high-performance kernels for attention, MoE, quantization.
• Upstream improvements to vLLM, SGLang, PyTorch; liaise with hardware and open-source communities.
• Shape hardware roadmap using roofline analysis and real workload data.
• Show passion for AI infrastructure and performance optimization.

Skills

C++

Python

GPU computing

Profiling

Kernel development

Education

Bachelors degree in CS/SE/AI

Tools

Triton

CUDA

PyTorch

vLLM

SGLang

Job description

Intel is seeking an AI Infrastructure Engineer in Santa Clara to push LLM inference on Intel GPUs, optimizing the end-to-end stack from profiling to kernel development. You will upstream improvements to open-source platforms like vLLM, SGLang, and PyTorch and collaborate with architecture teams to shape future GPU roadmaps.

The role requires strong C++/Python skills and 4+ years in GPU computing or HPC, with a passion for performance and scale-out inference across multi-node setups.