GPU Optimization Engineer - Multimodal AI (CUDA/Triton)

Luma; United States; On-site; USD 195,000 - 395,000; Full time

Job summary

Luma is seeking an expert in GPU and transformer optimization to accelerate multimodal models. You will profile and optimize GPU/CPU/accelerator code to maximize throughput and minimize latency while preserving model quality.

You will write high-performance PyTorch, Triton, and CUDA kernels, develop fused operations, and drive distributed multi-node deployment. Strong profiling skills and deep transformer knowledge are essential.

Qualifications

• Expert-level Triton/CUDA programming and GPU optimization.
• Strong PyTorch skills, including kernel development and custom operations.
• Proficiency with profiling tools (NVIDIA Nsight, torch profiler, custom tooling).
• Deep understanding of transformer architectures and attention mechanisms.

Responsibilities

• Profile and optimize GPU/CPU/accelerator code for max utilization and minimal latency.
• Write high-performance PyTorch, Triton, and CUDA, dropping to custom operations when needed.
• Develop fused kernels and leverage tensor cores and modern hardware features across platforms.
• Optimize model architectures and implementations for distributed multi-node production deployment.
• Build performance monitoring and analysis tools and automation.
• Research and implement cutting-edge optimization techniques for transformer models.

Skills

Triton/CUDA programming

PyTorch

Profiling tools

Transformer internals

Tools

PyTorch

Triton

CUDA

NVIDIA Nsight

Torch profiler

Job description

Luma is seeking an expert in GPU and transformer optimization to accelerate multimodal models. You will profile and optimize GPU/CPU/accelerator code to maximize throughput and minimize latency while preserving model quality.

You will write high-performance PyTorch, Triton, and CUDA kernels, develop fused operations, and drive distributed multi-node deployment. Strong profiling skills and deep transformer knowledge are essential.