5 days ago
Responsibilities
- Analyze latency, memory, power, and numerical correctness of ML inference on Apple hardware.
- Optimize ML model performance across Apple CPUs, GPUs, Neural Engine, system memory, and ML accelerators.
- Investigate inference behavior from the software stack and low-level drivers through hardware debugging.
- Create scripts, utilities, and debugging tools to extract, analyze, and report performance and power metrics.
- Collaborate with research, software engineering, hardware engineering, and product teams and lead technical discussions.
Requirements
- Experience with ML inference, quantization, and performance-versus-accuracy tradeoffs.
- Familiarity with ML architectures including large language models, diffusion models, and convolutional neural networks.
- Familiarity with operating systems, embedded systems, CPU/GPU hardware architectures, computer architecture, memory, compilers, and hardware drivers.
- Highly proficient in Python, C++, and shell scripting.
- Familiarity with Linux or macOS.
- Strong verbal and written communication skills, including presenting and leading discussions with larger groups.
- Preferred qualifications include a master’s or PhD in computer science or a related discipline.
- Preferred experience includes CoreML, MPS Graph, Metal Performance Shaders, MLX, TFLite, ONNX, ExecuTorch, PyTorch, TensorFlow, JAX, Xcode, Swift, Objective-C, MLIR, LLVM, or TVM.
About Apple
We’re a diverse collective of thinkers and doers, continually reimagining what’s possible to help us all do what we love in new ways. And the same innovation that goes into our products also applies to our practices — strengthening our commitment to leave the world better than we found it. This is where your work can make a difference in people’s lives. Including your own. Apple is an equal opportunity employer that is committed to inclusion and diversity. Visit apple.com/careers to learn more.