4 days ago
Strasbourg, FranceIntern
Responsibilities
- Explore automatic optimization and auto-tuning techniques for high-performance AI kernels.
- Investigate compiler techniques for generating and optimizing specialized implementations from models or intermediate representations.
- Develop profiling-guided methods for evaluating and selecting implementation strategies.
- Analyze tensor lifetimes and data dependencies to improve on-chip memory management and reuse.
- Study interactions among computation scheduling, data movement, memory usage, and model structure.
- Evaluate optimization techniques on Generative AI workloads and accelerator architectures.
- Develop experimental compiler or performance-analysis infrastructure and translate results into engineering recommendations.
Requirements
- Be a student in Computer Science, Computer Engineering, Electrical Engineering, or a related field.
- Have a strong interest in compilers and program optimization, machine-learning systems, high-performance computing, parallel programming, computer architecture, or Generative AI and Large Language Model inference.
- Experience with C++, Python, LLVM/MLIR, GPU or accelerator programming, or machine-learning frameworks is valuable but not required.
- Be curious, comfortable investigating open-ended technical problems, and interested in the interaction between machine-learning workloads, compiler transformations, memory systems, and hardware execution.
Benefits
- Six-month internship based in Strasbourg, France.
- Hands-on experience with compiler and performance-engineering problems for modern Generative AI workloads.
- Exposure to intermediate representations, code generation, accelerator execution, and quantitative performance analysis.
- Opportunity to build and evaluate a focused research prototype within an experienced compiler team.
Categories
About Qualcomm
Qualcomm is a public semiconductor company headquartered in San Diego, founded in 1985, that designs and sells wireless chipsets and platforms for mobile devices, automotive, IoT, and networking, notably the Snapdragon application processors and 5G modems. It also licenses a large portfolio of cellular patents to device makers, generating revenue alongside chip sales; its technology underpins many Android smartphones and emerging automotive and edge-compute systems, and it trades on NASDAQ as QCOM.