AI Performance Engineer (Cloud AI Engineering), Sr | Staff | Sr. Staff
Core
Convert, optimize, and deploy AI models (LLMs, VLMs, diffusion) for efficient inference on Qualcomm hardware, mapping next-gen workloads to current and future hardware designs.
Role type
Senior Staff AI Performance Engineer (Cloud AI Engineering)
Builds
Inference acceleration solutions for GenAI models
Domain
Cloud AI, Machine Learning, Hardware Acceleration
Deliverable
production ML models
Required skills
PyTorch, ONNX, Python, Transformer architectures, Attention mechanisms, Workload mapping (sharding, parallelisms), Computer architecture, ML accelerators, Distributed systems, Triton, Performance analysis
Preferred skills
Neural network operators, Machine learning compilers, Accuracy evaluation, torch.compile, torchDynamo
Technologies
PyTorch, ONNX, Triton, torch.compile, torchDynamo
Responsibilities
Convert, optimize, and deploy models for efficient inference; Identify optimization opportunities in advanced algorithms (attention, MoEs); Scale performance for throughput and latency constraints; Map AI workloads to hardware designs; Collaborate with compiler, firmware, and platform teams; Analyze performance/stability issues; Design and implement high-level kernels in Triton
Seniority
Senior Staff, hands-on IC
