Open-Source LLM GPU Frameworks Intern

Intel Corporation; Województwo pomorskie; Hybrid; PLN 62,000 - 76,000; Part time

Job summary

Intel Corporation in Poland is offering a hands-on AI internship focusing on building and optimizing the open software stack for LLM inference on Intel GPUs. You will work on vLLM and associated kernels, with mentorship from senior engineers.

The internship lasts 12–24 months and is part-time alongside university studies. As a student, you will write and test kernels, benchmark performance, and contribute to open-source projects used across the AI industry.

Qualifications

• Currently pursuing a BS or MS in Computer Science, Computer Engineering, AI, Mathematics, Physics, or related field.
• Strong Python programming skills.
• Understanding of deep learning and neural networks, incl. transformer architectures and LLM inference.
• Familiarity with Linux environments and strong analytical thinking.

Responsibilities

• Develop and optimize AI framework software for LLM inference on Intel GPUs under senior guidance.
• Implement, port, and tune compute kernels for core transformer building blocks.
• Run benchmarks, profile workloads, and report performance bottlenecks.
• Extend unit tests, correctness checks, and performance regression tests.
• Debug failures across the stack from Python to kernel hardware.
• Collaborate with global teams and participate in code reviews for open-source projects.

Skills

Python

Analytical thinking

Problem solving

Team collaboration

English communication

Education

Pursuing BS or MS in Computer Science/Engineering/AI/Math/Physics

Tools

PyTorch

Triton

CUDA

SYCL

OpenCL

vLLM

Job description

Intel Corporation in Poland is offering a hands-on AI internship focusing on building and optimizing the open software stack for LLM inference on Intel GPUs. You will work on vLLM and associated kernels, with mentorship from senior engineers.

The internship lasts 12–24 months and is part-time alongside university studies. As a student, you will write and test kernels, benchmark performance, and contribute to open-source projects used across the AI industry.