Junior Compiler Engineer (GPU Backend)

OXMIQ is developing a new class of GPU and AI silicon for large-scale model inference and training. The compiler stack connects industry-standard frameworks — PyTorch, JAX, ONNX, and Triton — to OXMIQ hardware, and is central to the performance and efficiency the company delivers to its customers. The Role The Junior Compiler Engineer will work under the direction of a senior compiler engineer to develop and extend the MLIR/LLVM-based compiler stack that targets OXMIQ hardware IP and GPU architectures. Responsibilities will be scoped to well-defined components of the toolchain — passes, lowerings, code generation features, and runtime hooks — with scope and autonomy expanding as proficiency develops. The use of AI-assisted development tools (Claude Code or equivalent) is a required part of the role and is expected to be applied throughout the engineer's daily work, including code authoring, review, debugging, and documentation. Key Responsibilities Develop and extend MLIR-based compilation passes targeting OXMIQ hardware IP and GPU architectures, under the technical direction of senior team members. Contribute to LLVM backend components, including instruction selection, register allocation, and target-specific optimizations. Implement code generation improvements that advance throughput, latency, and memory-efficiency objectives defined by the architecture and hardware teams. Collaborate with kernel engineers on Triton-based programming models and custom operator lowering. Participate in design and code reviews, performance profiling, and root-cause analysis of compiler-generated code. Develop tests, reproducers, and regression suites, and investigate correctness and performance regressions on production workloads. Required Qualifications Demonstrated proficiency with AI-assisted development tools (Claude Code, Cursor, or equivalent), with evidence of using such tools to deliver substantive production-quality code. Proficiency in modern C++ and working knowledge of Python. Working understanding of compiler intermediate representations, pass infrastructure, and code generation fundamentals. Familiarity with GPU programming models (CUDA, Triton, or comparable) or a demonstrated capacity to learn them. Proficiency in debugging and in standard Linux development tooling, including Git and CMake. Preferred Qualifications Exposure to LLVM or MLIR through coursework, internship experience, equivalent practical work, or contributions to LLVM, MLIR, Triton, or comparable open-source compiler projects. Experience with Triton, CUDA, or other GPU kernel programming frameworks. Familiarity with PyTorch, JAX, ONNX, or graph-level compilers such as XLA, TVM, IREE, or StableHLO. Exposure to GPU or AI accelerator backends, kernel fusion, quantization, or mixed-precision compilation. Education BS/MS/PhD in Computer Science, Computer Engineering, Electrical Engineering, or a related field — or equivalent practical experience. The successful candidate will work within a compact, senior team comprising compiler engineers, hardware architects, and kernel authors. A senior compiler engineer will serve as the primary technical mentor, providing direction, code review, and guidance across MLIR, LLVM backend development, and GPU code generation on production silicon. OXMIQ offers a competitive compensation package, including base salary, equity participation, comprehensive medical, dental, and vision coverage, and the opportunity to contribute to foundational silicon and software technology. OXMIQ is an equal opportunity employer. We evaluate qualified applicants without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, veteran status, or any other legally protected characteristic. #J-18808-Ljbffr Oxmiq Labs