Senior AI Performance Architect

Expired

Qualcomm

$126,700 – $217,900 · Posted 18 days ago

San DiegoLast seen 16 days ago

Apply →

Summary

Qualcomm seeks a Senior AI Performance Architect to design hardware accelerator architectures optimized for AI training at scale. The role involves analyzing GPU/accelerator architectures, defining computational blocks supporting multiple datatypes and precisions, architecting memory and scale-out systems, and codesigning hardware with software/LLM requirements. Responsibilities include competitive analysis, performance modeling, and pre-silicon prediction for ML training workloads. The ideal candidate spans software architecture, algorithm development, kernel optimization, and hardware accelerator design.

Data Science Embedded Systems AI C++Machine Learning Python Accelerator Architecture AI Training Systems Cnn Communication Protocols Compute In Memory Deepspeed Digital Circuits Distributed Training Fixed Point Floating Point Fsdp Gemm Gemv Gpu Architecture Hardware Modeling Hardware Simulators LLM Large Language Models Memory Technologies Microscaling Noc Network On Chip Nvidia Dgx Nvl72 Performance Modeling Power Analysis PyTorch Rnn Soc System On Chip Sparse Operations TensorFlow