Edge AI Deep Learning Compiler Engineer (Equity)

quadric technologies; Burlingame (CA); On-site; USD 170,000 - 230,000; Full time

Benefits offered by this job

Medical, dental, and vision insurance

Equity

401(k)

Paid Parental Leave

Catered lunch

Caltrain parking reimbursement

Office near Burlingame center

Job summary

Quadric is building the world’s first GPNPU and seeks a Deep Learning Compiler Engineer to design and optimize the compiler stack that translates modern neural networks to our hardware. You will influence performance, memory efficiency, and latency for edge AI workloads.

You’ll collaborate with hardware teams, implement graph optimizations, and push new operators across Transformer and Vision models while targeting FPGA/ASIC validation.

Qualifications

• BS/MS/PhD in Computer Science, Computer Engineering, Electrical Engineering, or related field.
• 8+ years hands-on experience developing deep learning compilers or compiler infrastructures (MLIR, TVM, LLVM, XLA, TensorRT, Glow).
• Strong proficiency in C++ (14/17/20) and Python, with solid fundamentals in data structures, algorithms, and OO design.
• Familiarity with modern AI/ML frameworks (PyTorch, TensorFlow, ONNX) and DL operator representations.
• Solid understanding of computer systems, memory hierarchies, parallel processing, and CPU/GPU/NPU execution.

Responsibilities

• Design, implement, and maintain compiler optimization passes targeting Quadric’s processor using MLIR, TVM, or LLVM.
• Develop lowering paths from high-level frameworks (PyTorch, TensorFlow, ONNX) to optimized low-level kernels.
• Optimize neural network performance for memory throughput, latency, and power consumption.
• Implement graph-level optimizations (fusion, layout transformation, quantization) and extend compiler support for new operators.
• Benchmark and profile model performance to identify and resolve bottlenecks.
• Collaborate with hardware teams to define ISA extensions and accelerator features.
• Develop simulators, models, and test benches for compiler correctness and performance.

Skills

C++

Python

MLIR

Apache TVM

LLVM

XLA

TensorRT

Glow

Education

Bachelor's degree

Master's degree

Ph.D.

Tools

PyTorch

TensorFlow

ONNX

Job description

Quadric is building the world’s first GPNPU and seeks a Deep Learning Compiler Engineer to design and optimize the compiler stack that translates modern neural networks to our hardware. You will influence performance, memory efficiency, and latency for edge AI workloads.

You’ll collaborate with hardware teams, implement graph optimizations, and push new operators across Transformer and Vision models while targeting FPGA/ASIC validation.