Senior/Staff Deep Learning Compiler...

About the Role

Join Quadric in redefining edge AI as a Deep Learning Compiler Engineer. You will be instrumental in designing and optimizing our compiler stack (utilizing MLIR, TVM, or LLVM) to bridge state-of-the-art neural network frameworks directly onto our proprietary GPNPU architecture.

Your Responsibilities

• Compiler Infrastructure & Optimization: Design and implement optimization passes for Quadric's processor. Develop lowering pathways from high-level ML frameworks (PyTorch, TensorFlow, ONNX) to optimized low-level kernel code. Optimize neural networks for memory throughput, latency, compute utilization, and power efficiency.
• Model Parsing & Graph Transformation: Implement graph-level optimizations such as operator fusion, layout transformation, quantization (INT8/FP16), and efficient memory allocation strategies. Analyze novel model topologies like Transformers and Vision-Language Models to extend compiler support for new operators.
• Hardware-Software Co-Design: Collaborate with hardware teams to define instruction set extensions and acceleration features. Develop simulators, functional models, and test benches to validate compiler correctness on FPGA and ASIC platforms before production bring-up.

Success Metrics

In your first 6–12 months, you will successfully integrate support for key deep learning architectures (e.g., Transformers), implement custom optimizations that deliver measurable performance improvements, and partner with architecture teams to influence next-generation micro-architecture designs through data-driven analysis.