AI Computing Architecture Researcher

Position Overview

We seek dedicated and skilled professionals to join a leading global technology company's team specializing in AI computing architecture research. This role involves designing and improving computing systems that support advanced AI workloads and intelligent agent functionalities.

Primary Responsibilities

• Architect and enhance AI computing platforms that support large language models (LLM), artificial intelligence-generated content (AIGC), distributed AI, and intelligent agents.
• Develop software components interfacing with hardware accelerators like GPUs, NPUs, and specialized AI processors.
• Advance the efficiency of distributed computing methods, parallel processing, and hardware accelerations.
• Work closely with AI researchers to optimize platform performance for intricate AI and agent-driven applications while addressing system-wide performance challenges.
• Create and maintain tools and frameworks for AI workload deployment, monitoring, scaling, and orchestration.
• Incorporate new AI models, agent controllers, and algorithms into platforms ensuring scalability, robustness, and efficient operation.
• Perform extensive profiling, benchmarking, and optimization to maximize throughput and minimize latency.
• Engage in research focusing on agent capabilities such as skill abstraction, composition, learning processes, and deployment mechanisms.
• Continuously update knowledge concerning AI hardware innovations, system architectures, and agent technologies to enhance platform functionality.

Required Qualifications

• Master's degree or PhD in Computer Science, Electrical Engineering, or closely related fields, with a preference for candidates holding a PhD and significant experience in high-performance AI computing systems.
• Strong programming skills in C/C++, Python, and AI domain-specific languages centered on low-level, high-performance computing tasks.
• Comprehensive understanding of AI model optimization methods including quantization, graph pruning, parameter compression, and sparsity algorithms.
• Solid knowledge in parallel programming, distributed system architectures, and multi-agent coordination techniques.
• Hands-on experience programming GPUs/NPUs using technologies such as CUDA, Triton, cuTile, or similar domain-specific languages.
• Proficiency with machine learning frameworks like PyTorch, TensorFlow, or MXNet.
• Familiarity with large-scale AI model deployment, orchestration frameworks, agent skill deployment, and runtime environment management.
• Experience utilizing containerization technologies including Docker and Kubernetes, and deploying large language models using platforms such as SGLang, vLLM, or HuggingFace along with cloud services like AWS, GCP, or Azure.
• Exceptional analytical and problem-solving capabilities with a focus on enhancing software performance for both traditional AI and agent-based systems.