On-device ML Performance Engineer, Graphics, Games and Machine Learning

What you'll do

• Analyze and optimize the performance, memory usage, and power consumption of ML models on Apple Silicon.
• Implement model conversion, compilation, and inference optimizations for various machine learning frameworks like PyTorch and MLX.
• Develop scripts and tools to extract, analyze, and report performance metrics for Apple hardware.
• Debug performance and power issues across CPU, GPU, and the Apple Neural Engine.
• Evaluate trade-offs between model accuracy and performance using techniques like quantization and sparsity.
• Provide data-driven insights on ML performance to internal and external stakeholders.
• Support the design and delivery of the on-device machine learning software stack.

What we're looking for

• Experience with ML inference, quantization, performance, and accuracy.
• Familiarity with popular ML architectures such as LLMs, Diffusion models, and CNNs.
• Knowledge of Operating Systems, embedded systems, and CPU/GPU/SoC/Memory hardware architectures.
• Proficiency in Python, C++, and shell scripting.
• Familiarity with Linux or macOS operating systems.
• Strong verbal and written communication skills to present data and lead discussions.
• Preferred: Master's or PhD in Computer Science or a related field.
• Preferred: Experience with Apple frameworks (CoreML, Metal), ML libraries (PyTorch, JAX), or compiler stacks (MLIR/LLVM).