GPU Kernel Performance Engineer — HPC/AI Infrastructure

Designworks Talent; Bellevue (WA); Hybrid; USD 140,000 - 190,000; Full time

Job summary

Designworks Talent in Bellevue, WA is seeking a HPC/GPU Performance / Kernel Engineer to optimize GPU kernels and boost AI workloads across training and inference.

You will work with CUDA/ROCm, profile performance, reduce latency, and collaborate with AI infrastructure and platform teams to push the limits of GPU efficiency.

This hybrid role offers ownership, advanced tooling, and the chance to influence architecture while supporting a growing AI platform.

Qualifications

• Bachelor’s or higher in Computer Science, Electrical Engineering, or related field.
• Strong experience with CUDA, ROCm, or equivalent GPU programming frameworks.
• Proven track record of optimizing GPU kernels and reducing latency in production AI workloads.

Responsibilities

• Profile, analyze, and optimize GPU kernels to improve latency and throughput.
• Identify data-plane bottlenecks across large AI workloads.
• Tune workloads for training and inference environments.
• Collaborate with AI infrastructure and platform engineering teams to optimize system behavior.

Skills

CUDA

ROCm

GPU kernel

Performance optimization

Systems programming

Distributed AI

Education

BS in CS/EE or related

Tools

Nsight Systems

Nsight Compute

ROCm profiling tools

GPU profiling tools

Job description

Designworks Talent in Bellevue, WA is seeking a HPC/GPU Performance / Kernel Engineer to optimize GPU kernels and boost AI workloads across training and inference.

You will work with CUDA/ROCm, profile performance, reduce latency, and collaborate with AI infrastructure and platform teams to push the limits of GPU efficiency.

This hybrid role offers ownership, advanced tooling, and the chance to influence architecture while supporting a growing AI platform.