GPU Engineer

Unikie is seeking a GPU Engineer with expertise in Linux media frameworks, GPU optimization, and embedded development. The successful candidate will enhance graphics and media pipelines for edge devices such as drones, laptops, and robotics platforms while utilizing NVIDIA, AMD, and Intel technologies. The role focuses on balancing real-time streaming performance with thermal and power efficiency and includes opportunities for open-source collaboration.

Key Responsibilities:

• Graphics/Media Pipeline Optimization: Design and optimize systems for real-time streaming and media processing on edge devices.
• GPU Performance Tuning: Improve performance for real-time and low-latency streaming using NVIDIA and other accelerators.
• Linux Driver & Kernel Development: Develop and maintain Linux drivers for NVIDIA, AMD, and Intel graphics hardware.
• Cross-Platform Acceleration: Optimize hardware acceleration for video encoding/decoding using CUDA, OpenGL, Vulkan, TensorRT, and GStreamer.
• Virtualization & Security: Use Linux virtualization technologies to securely isolate workloads and ensure robust performance.
• Troubleshooting & Performance Analysis: Diagnose and debug performance bottlenecks in media processing pipelines.
• Open Source Contributions: Contribute to Linux graphics and media subsystems projects.
• Documentation & Mentorship: Create technical documentation and mentor junior engineers.
• Applied R&D: Collaborate with researchers to develop functional prototypes from advanced concepts.

Required Qualifications:

• 5+ years of experience in graphics/media software development on Linux systems.
• Strong understanding of NVIDIA GPUs and familiarity with AMD and Intel platforms.
• Extensive experience with Linux kernel graphics and media subsystems, including custom driver development.
• Experience with GStreamer, V4L2, VAAPI, or similar pipelines.
• Proficiency with OpenGL, Vulkan, and real-time streaming protocols.
• Expertise in developing and debugging Linux device drivers for graphics hardware.
• Mastery of C/C++ with performance profiling and debugging tools.
• Proven track record of contributions to the Linux kernel, GPU drivers, or media frameworks that can be verified.

Selected candidates will be contacted for a discussion.