GPU Engineer Team Leader

ABOUT MOREH VIETNAM

Moreh Vietnam is developing Moreh-vLLM, a highly optimized inference framework for LLMs and generative models across diverse GPU and NPU architectures. Our innovations span GPU kernel development and optimization for single-GPU and multi-GPU systems, AI compiler optimization and R&D at the intersection of HPC and AI framework development.

ROLE OVERVIEW

We are looking for a GPU Engineer Team Leader to lead a focused team of 4–5 engineers in building and optimizing high-performance GPU software. You will own technical delivery, guide the team's day-to-day work, conduct code reviews, and grow each engineer's skills — while staying hands-on with the most critical and complex challenges for GPU.

This role sits at the intersection of technical excellence and team leadership. You are expected to be a strong individual contributor as well as a people-first leader who takes responsibility for the team's output, culture, and growth.

KEY RESPONSIBILITIES

Team Leadership & Delivery

• Lead, mentor, and develop a team of 4–5 GPU/HPC engineers, fostering a culture of technical excellence and continuous improvement.
• Plan and coordinate sprint tasks, manage workload distribution, and ensure timely, high-quality delivery of GPU software components.
• Conduct regular 1:1s, provide constructive feedback, and support engineers' career development.
• Translate high-level technical goals from senior engineers or engineering management into actionable team tasks.
• Track team progress, surface blockers early, and communicate status clearly to stakeholders.

Hands-on Technical Contribution

• Write production-quality GPU kernel code in CUDA, HIP, or OpenCL for AI training and inference workloads.
• Lead code reviews and enforce coding standards, GPU optimization patterns, and software quality across the team.
• Conduct performance profiling and optimization of GPU kernels and memory hierarchies
• Contribute directly to the most technically challenging features, debugging complex issues that require deep GPU systems knowledge.

MINIMUM QUALIFICATIONS

• Bachelor's degree in Computer Science, Computer Engineering, or a related field.
• Strong proficiency in C++ and Python.
• Experience with CUDA, HIP, or OpenCL.
• Experience with GPU memory hierarchy optimization (shared memory, registers, coalescing, occupancy).
• Familiarity with deep learning frameworks (PyTorch or TensorFlow) and how they interact with GPU computers.
• Demonstrated experience leading or mentoring a small engineering team (2+ people) in a technical setting.
• Strong analytical and problem-solving skills; ability to diagnose and resolve complex GPU software issues.
• Good written and verbal communication skills for team coordination and documentation.

PREFERRED QUALIFICATIONS

• Master's degree or Ph.D. in Computer Science, Computer Engineering, AI, or related field.
• 2+ years of experience writing system software for GPUs in a professional setting.
• Experience with distributed GPU computing, multi-GPU coordination, or parallel runtime systems.
• Knowledge of AI model architecture and its impact on GPU workload design (e.g., attention mechanisms, matrix operations).
• Track record of successfully owning delivery for a team or module end-to-end.
• Experience with profiling tools such as Nsight Compute, Nsight Systems, or AMD ROCm profiler.
• Contributions to open-source GPU/HPC projects or publications at relevant conferences (PPoPP, HPDC, SC, MICRO, etc.).