GPU Systems Engineer

GPU Systems Engineer - Remote 
 
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. 
 
Job Title: GPU Systems Engineer
Location: 100% Remote (U.S.) 
Position Type: Full-time, Direct W2 
Salary Range: $100,000–$150,000 Annually 
Experience Required: 6+ 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. 
 
Job Summary 
We are seeking a GPU Systems Engineer with deep expertise in CUDA programming, GPU architecture, and high-performance computing to design and optimize compute-intensive workloads on modern accelerator hardware. This role focuses on extracting maximum performance from GPU platforms for AI training, inference, scientific computing, and high-throughput data processing workloads. The ideal candidate combines low-level systems mastery with strong software engineering practices, and has a track record of delivering measurable performance improvements on production GPU systems. In this role you will work closely with cross-functional partners — product, design, engineering, operations, and business stakeholders — to translate ambiguous requirements into well-engineered solutions, and will be expected to raise the bar through code review, design review, and mentorship of more junior engineers. The successful candidate brings strong engineering discipline, a clear communication style, and a track record of shipping meaningful work that holds up well in production. 
Key Responsibilities 

• Design and implement high-performance CUDA kernels for compute-intensive workloads across AI and HPC use cases. 
• Profile and optimize GPU code using tools such as Nsight Systems, Nsight Compute, and CUDA profilers. 
• Tune memory access patterns, occupancy, register usage, and shared memory utilization for peak performance. 
• Develop highly optimized libraries for linear algebra, attention, and other ML primitives. 
• Optimize multi-GPU and multi-node training using NCCL, RDMA, and high-performance networking. 
• Implement custom operators and fused kernels in PyTorch, JAX, or Triton. 
• Collaborate with ML engineers to identify performance bottlenecks in training and inference pipelines. 
• Develop benchmarks and regression tests to safeguard performance over time. 
• Evaluate new GPU architectures and feature sets, and advise on adoption strategy. 
• Contribute to compiler-level optimizations for tensor programs where appropriate, working at the boundary between ML frameworks and underlying accelerator codegen to unlock performance not reachable through framework-level tuning alone. 
• Optimize memory hierarchy usage across HBM, L2, shared memory, and registers. 
• Implement mixed-precision and quantized compute paths that maximize accelerator throughput while preserving numerical fidelity within bounds acceptable for the target workloads. 
• Document performance characteristics, design decisions, and tuning playbooks for internal teams. 
• Stay current with GPU architecture, CUDA evolution, and emerging accelerator technologies. 

Required Qualifications 

• Bachelor’s or Master’s degree in Computer Science, Computer Engineering, or a related field. 
• Six or more years of experience in GPU programming and performance engineering. 
• Deep expertise in CUDA C/C++ and GPU programming models. 
• Strong understanding of modern GPU architectures, memory hierarchies, and execution models. 
• Hands-on experience profiling and optimizing GPU workloads in production. 
• Familiarity with NCCL, MPI, and high-performance interconnect technologies. 
• Experience integrating custom kernels into ML frameworks. 
• Strong C++ skills and familiarity with modern systems programming practices. 
• Solid grounding in linear algebra and numerical methods. 
• Strong communication and collaboration skills with research and engineering teams. 

Preferred Qualifications 

• Experience with Triton, CUTLASS, or other GPU kernel authoring frameworks. 
• Familiarity with TensorRT, FasterTransformer, or vLLM internals. 
• Exposure to compiler infrastructure such as LLVM or MLIR. 
• Open-source contributions to GPU or ML performance libraries. 
• Experience with large-scale distributed training infrastructure. 

How to Apply 
Would you like to know more about this opportunity? For immediate consideration, please send your resume to [email protected] or contact us at (908) 505-3544. Learn more about Bright Vision Technologies at www.bvteck.com. 
Bright Vision Technologies is an Equal Opportunity Employer.