ML Algorithm Mapping and Performance Engineer, Core ML

What you'll do

• Build performance models, benchmarks, and prototypes; analyze algorithm behavior as workloads and hardware scale.
• Assess performance trade-offs and bottlenecks across quality, latency, throughput, memory, communication, and compute.
• Collaborate across research and engineering teams, and communicate recommendations through reports, presentations, and design reviews.

What they ask for

• Bachelor’s, Master’s, PhD, or equivalent practical experience in a related field.
• Strong foundation in computer architecture, parallel computing, systems performance, and machine learning systems.
• Experience with analytical performance modeling, complexity analysis, benchmarking, or system simulation; profiling and debugging performance in a relevant system.
• Proficiency in Python and comfort with C++.

Nice to have

• Experience with roofline analysis, simulators, kernel optimization, or hardware–software co-design.
• Familiarity with CUDA, Triton, PyTorch, JAX, or open-source LLM systems.

Tools

Python, C++, Cerebras WSE, CUDA, Triton, PyTorch, JAX

AI focus

Develops and maps efficient ML training and inference algorithms, including techniques such as parallel token generation, diffusion and speculative decoding, attention, sparsity, mixture-of-experts, low-precision computation, and distributed training.