Software Engineer, Planner GPU Compute

Job overview

The Software Engineer, Planner GPU Compute role is onsite in Boston, MA, with a base salary range of $209,000–$263,000 per year. Zoox builds purpose-built robotaxis, and this position supports the Planner Compute team responsible for motion-planner performance in the autonomy stack. The role exists to improve GPU performance so the motion planner can operate safely with low latency and consistent resource utilization. The engineer analyzes GPU performance, adapts the planner to different GPU architectures, optimizes machine-learning models, and advises Planner engineers on CUDA and GPU performance.

What you'll do

• Analyze GPU performance metrics to identify hotspots and optimizations
• adapt the planner to multiple GPU architectures
• optimize machine-learning models for latency and GPU memory use
• advise Planner engineers on CUDA and GPU performance.

What we're looking for

Skills & competencies

software engineer

planner gpu compute

gpu performance

cuda

c++

motion planner

autonomy

boston ma

zoox

$209k–$263k/year

Work arrangement

• Weekend coverage required

Benefits & perks

• Paid time off, health insurance, long-term care insurance, disability insurance, life insurance, Amazon RSUs, and Zoox Stock Appreciation Rights.

Why this role

The role spans GPU use from machine-learning inference to custom CUDA kernels for the motion planner.

About the employer

Zoox is hiring for this role. Industry: Custom Computer Programming Services. Sector: 54.