Remote CUDA Kernel Engineer - Optimize AI GPU Pipelines

Pragmatike; Town of Florida (NY); On-site; USD 120,000 - 150,000; Full time

Benefits offered by this job

Competitive salary & equity options

Sign-on bonus

Health, Dental, and Vision

401(k)

Job summary

Pragmatike is seeking a CUDA Kernel Engineer to develop and optimize NVIDIA CUDA kernels for high-throughput AI systems. This role involves designing custom kernels, profiling GPU workloads, and resolving performance bottlenecks. Ideal candidates will have a strong understanding of NVIDIA GPU architectures and experience in optimizing GPU performance.

As part of a fast-growing AI startup, you'll directly impact client solutions for Fortune 500 companies. Benefits include competitive salary, equity options, and comprehensive health benefits.

Qualifications

• Proven track record building NVIDIA CUDA kernels from scratch.
• Strong ability to optimize kernels, including tiling strategies and warp scheduling.
• Deep understanding of GPU memory hierarchy and performance diagnostics.

Responsibilities

• Design and optimize custom CUDA kernels for NVIDIA GPUs.
• Profile GPU workloads and analyze performance bottlenecks.
• Collaborate with AI systems and backend distributed teams.

Skills

CUDA kernel optimization

NVIDIA GPU architecture understanding

Profiling workflows

C++

Tools

Nsight Compute

CUDA-MEMCHECK

Nvprof

Job description

Pragmatike is seeking a CUDA Kernel Engineer to develop and optimize NVIDIA CUDA kernels for high-throughput AI systems. This role involves designing custom kernels, profiling GPU workloads, and resolving performance bottlenecks. Ideal candidates will have a strong understanding of NVIDIA GPU architectures and experience in optimizing GPU performance.

As part of a fast-growing AI startup, you'll directly impact client solutions for Fortune 500 companies. Benefits include competitive salary, equity options, and comprehensive health benefits.