Intel · posted · checked
Senior Compiler Performance Engineer
C++Pytorch$156k – $220kHybrid · TorontoSenior
Who can apply
Hybrid · Toronto
Location: Canada, Toronto
Job Details:
Job Description:
The Compiler Engineering team within Intel's Data Center Group is seeking talented and passionate Senior Compiler Engineers to help shape the future of high-performance computing and AI software.
At Intel, we are developing next-generation software technologies that unlock the full potential of heterogeneous computing. Our compiler technologies enable developers to efficiently leverage CPUs, GPUs, and accelerators through high-performance, portable software. As part of our team, you will contribute to an ecosystem that advances performance, scalability, reliability, and developer productivity for AI, HPC, and other demanding workloads, while supporting Intel's commitment to open-source innovation and collaboration.
The Intel Compiler Engineering team develops LLVM-based compiler technologies for C, C++, SYCL, Triton, MLIR, Fortran, and Domain-Specific Languages (DSLs), targeting Intel CPU and GPU architectures. We work across the entire compiler stack and collaborate closely with hardware architecture teams, covering language front ends, intermediate representations, optimization, code generation, debug information, runtimes, heterogeneous and offload computing, quality engineering, and developer tools. Our work bridges Intel hardware capabilities with leading AI frameworks and software ecosystems, including PyTorch, machine learning frameworks, and inference platforms.
Responsibilities
As a Senior Compiler Engineer, you will play a key role in advancing compiler technologies that deliver industry-leading performance and scalability across Intel architectures. You will take ownership of complex technical challenges from concept to production and drive solutions that have a broad impact across products and workloads.
In this role, you will:
Lead performance analysis of real-world applications, AI and HPC workloads, and industry-standard benchmarks, identifying bottlenecks across compilers, runtimes, libraries, and processor architectures.
Design, develop, and deliver compiler optimizations and code generation enhancements that enable new Intel CPU and GPU capabilities, taking innovations from prototype through production-quality implementation.
Own performance for specific compiler components or workload domains by defining targets, tracking results through continuous integration systems, and driving measurable improvements.
Advance heterogeneous and offload compilation technologies across SYCL, OpenMP, Triton, MLIR, and DSL-based programming models, improving portability while maintaining exceptional performance.