AI Software Engineer, SoC Tooling
Analog Devices Vezi toate joburile
- Cluj-Napoca, Cluj
- Permanent
- Full-time
- Design and evolve intermediate representations that describe computation, memory movement, and execution across multiple cores.
- Implement transformation pipelines that progressively refine high-level model descriptions into hardware-executable forms.
- Develop approaches to map computations efficiently onto heterogeneous compute resources.
- Define compilation passes that transform generic AI models into optimized representations suitable for efficient hardware execution.
- Develop model optimization strategies such as graph simplification, operator fusion, quantization, layout transformation, and precision scaling.
- Optimize for latency, throughput, and energy efficiency on edge and embedded systems.
- BS/MS in Electrical Engineering (EE), Computer Engineering (CE) or Computer Science (CS) with minimum of 4 years of relevant experience in the design, implementation, and support of AI software, compiler infrastructure, or systems software development.
- Strong software development skills in multiple programming languages, including C/C++ and Python.
- Experience with embedded systems
- Experience with SCM (svn, git)
- Excellent software design, problem solving and analytical debugging skills
- Knowledge of machine learning model architectures and inference workflows.
- Familiarity with heterogeneous computing and memory hierarchies.
- Understanding of compiler concepts (IR design, lowering, optimization passes).
- Experience with embedded or edge AI systems.
- Familiarity with Ai model interchange formats and compiler frontends such as ONNX, StableHLO, Torch-MLIR or other similar ecosystems.
- Knowledge of real-time systems or power-aware computing.
- Experience working with compiler infrastructures such as LLVM, TVM, IREE or XLA.
- Experience with graph partitioning, multi-device scheduling, or heterogeneous execution.