Remote GPU Software Engineer (CUDA) — High-Performance AI & HPC

Bright Vision Technologies; Canton Charter Township (MI); On-site; USD 80,000 - 107,000; Full time

Job summary

Bright Vision Technologies seeks a GPU Software Engineer (CUDA) for 100% remote US work to optimize compute workloads on modern GPUs for AI training, inference, and data processing.

Ideal candidate has 7+ years in GPU programming, deep CUDA C/C++, and production profiling, and will collaborate with product, design, and engineering teams to deliver performant software.

Requirements include NCCL/MPI experience, kernel integration into ML frameworks, strong C++ skills, and a CS/CE degree (BS/MS).

Qualifications

• 7+ years of GPU programming and performance engineering.
• Strong CUDA C/C++ and GPU architectures knowledge.
• Experience profiling, debugging, and optimizing GPU workloads.

Responsibilities

• Design and optimize compute-intensive workloads on GPUs.
• Collaborate with cross-functional teams to translate requirements into robust software.
• Mentor junior engineers and participate in peer reviews.

Skills

CUDA C/C++

GPU programming

Performance profiling

MPI/NCCL

C++

High-performance computing

Communication

Education

CS/CE degree (BS/MS)

Tools

NCCL

MPI

Triton

CUTLASS

Job description

Bright Vision Technologies seeks a GPU Software Engineer (CUDA) for 100% remote US work to optimize compute workloads on modern GPUs for AI training, inference, and data processing.

Ideal candidate has 7+ years in GPU programming, deep CUDA C/C++, and production profiling, and will collaborate with product, design, and engineering teams to deliver performant software.

Requirements include NCCL/MPI experience, kernel integration into ML frameworks, strong C++ skills, and a CS/CE degree (BS/MS).