Triton Compiler Engineering Intern

Intel Corporation

2d • Remote

About The Position

Triton-lang is a open-source programming language and compiler for creating highly efficient custom GPU kernels. Its widely adopted in several stacks like Pytorch, SGLang, vLLM etc. Come join us to learn and contribute to Triton front end and back end components for Intel GPUs.

Requirements

• Candidate must be pursuing a Masters or PhD degree in Electrical Engineering, Electronics and Communication Engineering, Computer Engineering, Computer Science or a related field of study.
• 3+ months of experience with the following:
• Algorithms, data structures, and software development in C/C++.
• Compiler stages, code generation, and optimization techniques.
• GitHub development environment.

Nice To Haves

• Strong problem solving and analytical skills.
• Experience with LLVM, MLIR, and Triton compiler architecture.
• Familiarity with PyTorch attention techniques for transformer models.

Responsibilities

• Participate to define, design, develop, test, debug, release, enhance, and maintain software tools for domain-specific programming languages.
• Work directly with hardware design teams, compiler development companies, and programming communities.
• Design and implement Triton compiler front-end and back-end components.
• Develop optimization passes for supported kernels.

What This Job Offers

• Career Level: Intern
• Number of Employees: 5,001-10,000 employees