3 days ago
Shanghai, China or Beijing, ChinaIntern
Responsibilities
- Design, implement, and maintain optimized communication runtimes for deep learning frameworks and HPC programming interfaces on GPU clusters.
- Develop system software enabling communication among GPUs and between GPUs and other system components.
- Contribute to parallel programming interface specifications including MPI and OpenSHMEM.
- Create proof-of-concepts for programming-model extensions, runtime designs, and hardware features.
Requirements
- Pursuing a Ph.D. in computer engineering, computer science, or electrical engineering with a strong background in computer architecture, operating systems, communication libraries, or AI/ML.
- Excellent C/C++ programming and debugging skills.
- Strong experience with Linux, parallel programming interfaces, and communication runtimes.
- Ability to work and communicate effectively in a multinational, multi-time-zone corporate environment.
- Preferred qualifications include knowledge of InfiniBand and RoCE, HPC applications, PyTorch, TensorFlow, JAX/XLA, vLLM, SGLang, AI/DL communication patterns, CUDA kernel optimization and profiling, and large-scale model training or production inference software.
- Preferred qualifications include strong collaboration and interpersonal skills, including the ability to guide and influence in a dynamic matrix environment.
About Nvidia
Nvidia designs and sells GPUs and accelerated computing platforms for data centers, AI/ML, graphics, gaming, and automotive, monetizing through hardware, software platforms (CUDA, AI frameworks), and systems like DGX and networking. Customers include cloud providers, enterprises, researchers, and OEMs. Founded in 1993 and headquartered in Santa Clara, it is a public company traded on NASDAQ under NVDA.