AI Computing Architect - High-Performance Systems

Beijing Foreign Enterprise Management Consultants Co.,Ltd.; Singapore; Presencial; SGD 150.000 - 230.000; Jornada completa

Descripción de la vacante

Huawei in Singapore seeks an AI Computing Architecture Researcher to design and optimize high‑performance platforms for LLM and agent workloads. You will build software that interfaces with GPUs, NPUs and AI chips, and collaborate with researchers to boost throughput and reduce latency.

The ideal candidate has a PhD or Master’s in CS/EE, strong CUDA and parallel/distributed systems experience, and proficiency in PyTorch or TensorFlow.

Formación

• Advanced knowledge of high-performance AI systems and computing platforms.
• Expertise in low-level programming and performance optimization.
• Experience with AI model optimization techniques and multi-agent workloads.

Responsabilidades

• Design and optimize AI computing architectures for LLM, AIGC, distributed AI, and agent workloads.
• Develop software interfacing with GPUs, NPUs, and AI chips.
• Improve performance of distributed computing and GPU/NPU acceleration.
• Collaborate with AI researchers to enhance platform performance and resolve system bottlenecks.
• Create tools for deployment, monitoring, scaling, and orchestration of AI workloads.
• Integrate AI models and agents into scalable architectures with fault tolerance.
• Perform profiling and optimization for high throughput and low latency.
• Research agent capabilities including skill abstraction and deployment.

Conocimientos

C/C++

Python

GPU/NPU programming

Parallel programming

Distributed systems

Multi-agent coordination

AI frameworks (PyTorch/TensorFlow)

Containerization (Docker/Kubernetes)

Cloud infrastructure (AWS/GCP/Azure)

Performance optimization

Educación

PhD in CS/EE

Master's degree in CS/EE

Herramientas

CUDA

Triton

cuTile DSL

SGLang

HuggingFace

PyTorch

TensorFlow

Docker

Kubernetes

Cloud platforms (AWS/GCP/Azure)

Descripción del empleo

Huawei in Singapore seeks an AI Computing Architecture Researcher to design and optimize high‑performance platforms for LLM and agent workloads. You will build software that interfaces with GPUs, NPUs and AI chips, and collaborate with researchers to boost throughput and reduce latency.

The ideal candidate has a PhD or Master’s in CS/EE, strong CUDA and parallel/distributed systems experience, and proficiency in PyTorch or TensorFlow.