GPU Kernel Performance Engineer: Optimize AI Workloads

Acceler8 Talent; Mountain View (CA); On-site; USD 180,000 - 260,000; Full time

Job summary

Acceler8 Talent in Mountain View, CA is seeking a GPU / Kernel Performance Engineer to optimize performance, power efficiency, and cost for large-scale AI workloads. This onsite role offers deep ownership of kernel development and system-level improvements working with a small, technical team.

The candidate will profile workloads, optimize kernels, analyze GPU architectures, and collaborate with ML infrastructure and engineering teams to translate kernel-level gains into end-to-end performance

Qualifications

• Hands-on experience developing, modifying, or optimizing GPU kernels using CUDA or comparable GPU/HPC programming technologies.
• Strong understanding of GPU architecture, including memory hierarchies, parallel execution, synchronization, and hardware utilization.
• Ability to profile GPU workloads and identify performance bottlenecks.
• Experience implementing performance improvements and measuring their impact using benchmarks or profiling tools.
• Strong programming fundamentals and debugging low-level performance issues.
• Willingness to work across GPU kernels, compiler/runtime behavior, and broader software systems.
• Strong technical problem-solving skills and the ability to work effectively in a fast-moving engineering environment.

Responsibilities

• Profile GPU workloads to identify bottlenecks and efficiency opportunities.
• Optimize existing kernels and develop new ones to improve performance.
• Analyze architecture, memory, parallel execution, synchronization, and utilization.
• Investigate memory access, kernel fusion, occupancy, and data movement.
• Develop reproducible benchmarks and measure improvements on AI workloads.
• Validate numerical correctness and performance consistency after optimization.
• Collaborate with ML infrastructure, serving, and ML teams to translate kernel gains into system-wide performance.
• Investigate performance interactions across kernels, compilers, runtimes, and software.
• Evaluate communication and memory bandwidth trade-offs across multi-GPU systems.

Skills

CUDA kernel development

GPU architecture

GPU profiling

Benchmarking

GPU performance analysis

Programming fundamentals

Kernel/compiler collaboration

Problem solving

Tools

CUDA

Triton

CUTLASS

cuDNN

Nsight Systems

Nsight Compute

Job description

Acceler8 Talent in Mountain View, CA is seeking a GPU / Kernel Performance Engineer to optimize performance, power efficiency, and cost for large-scale AI workloads. This onsite role offers deep ownership of kernel development and system-level improvements working with a small, technical team.

The candidate will profile workloads, optimize kernels, analyze GPU architectures, and collaborate with ML infrastructure and engineering teams to translate kernel-level gains into end-to-end performance