Parallel Computing Engineer

Bright Vision Technologies is a technology consulting and software development company delivering cloud, AI, data, and enterprise solutions across the United States. This is a fantastic opportunity to join an established and well-respected organization offering tremendous career growth potential.

Parallel Computing Engineer

Location: 100% Remote (U.S.)
Position Type: Full-time, Direct W2
Salary Range: $130,000–$180,000 Annually
Experience Required: 10+ Years

Sponsorship: U.S. Citizens, Green Card Holders, EAD Holders, and H-1B transfer candidates are encouraged to apply. We are unable to sponsor new H-1B visa petitions for this position.

Bright Vision Technologies is seeking a highly experienced Parallel Computing Engineer with 10+ years of experience in High-Performance Computing (HPC), GPU programming, and parallel computing to optimize AI, machine learning, and scientific computing workloads. The ideal candidate will possess deep expertise in CUDA, GPU architecture, distributed computing, performance optimization, and large-scale AI infrastructure, with a proven track record of designing high-performance computing solutions for enterprise and research environments.

• Design, develop, and optimize high-performance CUDA kernels for AI, deep learning, and scientific computing applications.
• Analyze, profile, and optimize GPU workloads using NVIDIA Nsight Systems, Nsight Compute, CUDA Profiler, and related performance analysis tools.
• Optimize GPU memory management, kernel execution, multi-GPU scaling, and distributed computing performance.
• Design scalable distributed training and inference architectures using NCCL, MPI, CUDA-aware communication libraries, and high-performance networking technologies.
• Develop custom GPU operators and optimized kernels for PyTorch, JAX, Triton, TensorFlow, or similar AI frameworks.
• Improve training and inference performance for large language models (LLMs), deep learning, and high-performance AI workloads.
• Collaborate with AI researchers, ML engineers, and software architects to accelerate production AI applications.
• Build automated benchmarking frameworks, performance regression testing, and optimization pipelines.
• Evaluate emerging GPU technologies, programming models, and accelerator architectures to improve computational efficiency.
• Mentor engineers and provide technical leadership in GPU optimization, HPC architecture, and parallel programming best practices.

• Bachelor's or Master's degree in Computer Science, Computer Engineering, Electrical Engineering, or a related technical discipline.
• 10+ years of professional experience in GPU programming, High-Performance Computing (HPC), or parallel computing.
• Expert-level proficiency in CUDA C/C++, GPU architecture, and massively parallel programming techniques.
• Extensive experience with NCCL, MPI, CUDA-aware MPI, and distributed GPU communication frameworks.
• Strong understanding of GPU memory hierarchy, kernel optimization, occupancy tuning, and performance analysis.
• Hands-on experience integrating custom GPU kernels into PyTorch, TensorFlow, JAX, Triton, or other machine learning frameworks.
• Strong C/C++ programming skills with expertise in debugging, profiling, and performance optimization.
• Experience developing scalable AI or HPC solutions on cloud platforms or large GPU clusters.
• Excellent analytical, communication, collaboration, and technical leadership skills.

• Experience with Triton, CUTLASS, TensorRT, FasterTransformer, vLLM, DeepSpeed, or similar GPU optimization frameworks.
• Knowledge of LLVM, MLIR, compiler optimization techniques, or code generation technologies.
• Experience with large-scale distributed AI training, model parallelism, pipeline parallelism, and inference optimization.
• Familiarity with cloud-based GPU infrastructure on AWS, Microsoft Azure, or Google Cloud Platform (GCP).
• Contributions to open-source GPU libraries, research publications, patents, or technical presentations.
• Experience with emerging accelerator technologies such as AMD ROCm, Intel