Real-Time GPU Performance Engineer (C++, CUDA)

Front Door Defense

Foster City (CA)

On-site

USD 217,000 - 307,000

Full time

14 days+

Get more replies from employers

Send a job-specific resume in minutes.

Job summary

Zoox is seeking a Senior Software Engineer - C++ GPU Performance to optimize real-time autonomous driving workloads across Zoox compute platforms. You will instrument, monitor, and optimize GPU-based algorithms, from computer vision to multi-agent decision making, to maximize throughput and minimize latency.

You will build real-time instrumentation, analyze GPU hotspots, and help bring serial algorithms to the GPU.

Qualifications

BS in computer science or related field and 7+ years of experience.
Strong knowledge of CUDA on recent GPU architectures (e.g., Ampere, Blackwell) and Nsight-based debugging.
Strong knowledge of C++ in large code bases; comfortable with Linux development.
Experience in development, debugging, and profiling of complex multiprocess systems.

Responsibilities

Build real-time instrumentation for performance monitoring (CPU, GPU, latency, memory) and develop offline benchmarking frameworks and scripts to evaluate performance at scale.
Analyze performance metrics to identify GPU hotspots and propose actionable solutions with component teams.
Support teams in porting serial algorithms to the GPU to maximize compute utilization and reduce latency.
Design a middleware framework with the Core team to promote efficient CPU/GPU code development.