Senior AI Accelerator Architect - On-Device & Edge

MaC VC; San Diego (CA); On-site; USD 127,000 - 218,000; Full time

Job summary

Qualcomm Technologies, Inc. is seeking an AI Accelerator Architecture Engineer to drive architectural and performance improvements for AI training systems.

You will work on HW/SW co-design, data-paths, memory subsystems, and scalable architectures to support large-scale accelerators and data centers. You will evaluate current GPU/accelerator architectures, engage with customers to capture requirements, and help define models to prove architecture value.

Qualifications

• Requires advanced degree in CS/Engineering or related field.
• 3+ years hardware engineering experience defining GPU/accelerator architectures for AI training.
• Knowledge of NVIDIA/AMD GPU architectures and capabilities.
• Familiarity with LLM architectures and HW requirements.

Responsibilities

• Analyze ML network trends and determine SW/HW impact.
• Define hardware requirements for AI training systems with customers.
• Evaluate accelerator and GPU architectures and propose enhancements.
• Architect flexible compute blocks supporting multiple data types and precisions.
• Design memory subsystems optimized for AI workloads and compute-in-memory concepts.

Skills

C++

Python

HW/SW co-design

LLM hardware requirements

Strong communication

Education

Master's degree in CS/Engineering/Info Systems

Bachelor's degree in CS/Engineering

PhD (preferred)

Tools

NVIDIA DGX/NVL72

DeepSpeed (FSDP)

NoC/SoC design tools

Job description

Qualcomm Technologies, Inc. is seeking an AI Accelerator Architecture Engineer to drive architectural and performance improvements for AI training systems.

You will work on HW/SW co-design, data-paths, memory subsystems, and scalable architectures to support large-scale accelerators and data centers. You will evaluate current GPU/accelerator architectures, engage with customers to capture requirements, and help define models to prove architecture value.