AI Systems, Model Optimization

As a Member of Technical Staff, AI Systems, Model Optimization, you will develop the path from model architecture to physical silicon. You will develop the training techniques, optimization strategies, and infrastructure required to make AI models run efficiently on our novel compute substrates, closing the loop between model design and tapeout.

• Energy Benchmarking & Performance Modeling: Develop rigorous performance models to evaluate compute, memory, and energy trade-offs. Track pareto-optimality across models and hardware configurations.
• Advanced Mapping & Partitioning: Drive the partitioning and mapping of complex AI models down to hardware. Apply and invent advanced optimization strategies from first principles, including custom quantization schemes, sparsity/pruning, and distillation to fit the physical constraints of our substrates.
• Hardware-Aware Training: Develop and apply Quantization-Aware Training (QAT), noise-aware training, and sparsification techniques to adapt models to the physical constraints of our analog compute substrates, including memory footprint, connectivity, precision, and noise.
• GPU Optimization & Kernel Development: Develop and optimize kernels using low-level programming models like CUDA, Triton, or CUTLASS. Profile and debug complex ML codebases to resolve performance bottlenecks (training and inference).
• Cross-Functional Collaboration: Act as a translator between AI model architects and hardware/infrastructure engineering teams, converting model requirements into concrete specifications and codifying learnings for tapeouts.