Deep Learning Compiler Engineer (New Grad)
Core
Design and implement compiler passes to lower ONNX models through Relay IR to C++ code targeting the Chimera GPNPU for edge and automotive AI inference.
Role type
New-grad Deep Learning Compiler Engineer
Builds
Neural network inference software for Quadric GPNPU silicon
Domain
AI hardware acceleration / Compiler engineering
Deliverable
production ML models
Required skills
Python, C++, compiler intermediate representations, dataflow analysis, transformation passes, lowering, debugging large codebases
Preferred skills
null
Technologies
ONNX, Relay IR, C++, Python
Responsibilities
Own compiler passes, design and implement IR transformations, maintain pipeline segments, debug end-to-end, diagnose compilation issues via tracing, improve compiler decisions, partner with kernel and hardware teams, strengthen toolchain, contribute to test infrastructure
Seniority
New-grad, hands-on IC
