Research Engineer - Next-Generation Computing Architecture

About the Role

Our client, a global leader in information and communications technology infrastructure and smart devices, provides comprehensive AI solutions for carriers, enterprises, governments, and consumers worldwide. They are seeking a Research Engineer to join their Singapore research team focused on next-generation computing architectures. The role involves pioneering system and architecture research tailored to emerging AI workloads, such as efficient inference for agentic tasks, heterogeneous resource aggregation, memory tiering, and hardware-software co-design.

Key Responsibilities

• Lead research on systems and architecture targeting LLM (Large Language Model) and agentic workload challenges, covering inference serving, task scheduling, KV cache and state management, plus memory and storage hierarchies.
• Analyze authentic workloads and develop performance models alongside simulation frameworks to inform architectural design decisions.
• Create and deploy system prototypes, performing thorough end-to-end evaluations to verify technical innovations.
• Engage in hardware-software co-design efforts, collaborating closely with chip, compiler, and system software teams to enable practical implementations.
• Maintain awareness of advancements in the field, contribute to collaborative research projects with universities, and publish findings at premier academic conferences.

Qualifications & Requirements

• Possess a Master's degree or PhD in Computer Architecture, Computer Science, Operating Systems, Distributed Systems, Compilers, EDA, or closely related disciplines.
• Demonstrate solid expertise in at least one of the following areas: large-model inference and training systems, GPU/NPU accelerator architecture, memory and storage systems, high-performance networking and interconnects, compilers and runtimes, or chip design and verification.
• Exhibit strong programming and engineering capabilities with C/C++ and Python, supplemented by experience in CUDA or similar toolchains, with a track record of driving projects from prototype phase to performance testing.
• Have an established publication history in elite conferences such as ISCA, MICRO, ASPLOS, HPCA, OSDI, SOSP, NSDI, EuroSys, MLSys, or DAC.
• Possess excellent communication and teamwork skills, enabling productive collaboration across diverse teams and academic partners.

Additional Information

Selected candidates will collaborate extensively with researchers at local universities and industry partners. Please note that only shortlisted applicants will be contacted for further assessment.

EA License No.: 01C4394

By submitting your application, you agree to allow PERSOL Singapore Pte Ltd and its affiliates to collect, use, and disclose your personal data according to their Privacy Policy.