1 year ago
Shanghai, ChinaMid Level
Responsibilities
- Design, implement, and optimize GPU computing kernels for 3D generative AI model training and inference.
- Develop and maintain domain-specific libraries and performance-critical components for 3D generation workloads.
- Collaborate with researchers and infrastructure engineers to identify bottlenecks, benchmark performance, and deliver production-ready GPU modules.
Requirements
- Hands-on experience with CUDA and GPU programming.
- Strong programming skills in C++ and Python.
- Solid understanding of parallel programming, performance tuning, and numerical computation.
- Preferred experience with quantization, model compression, or other model optimization techniques.
- Preferred knowledge of computer graphics, rendering pipelines, or geometry processing.
- Preferred familiarity with GPU profiling tools such as Nsight and nvprof, or with hardware-aware optimization.
Benefits
- The posting describes a global team culture focused on knowledge, empathy, direct communication, innovation, quality, and aesthetics.
Categories
