Nanyang Technological University

Research Associate (GPU System Engineer)

The Rapid-Rich Object SEarch (ROSE) Lab is a joint research initiative between the Nanyang Technological University (NTU) and Peking University (PKU).  It aims to create a platform for fast mobile searches on object databases.  The three key thrusts of the ROSE Lab are:  (i) large-scale object database collection and analytics, (ii) scalable mobile object search with contextual mobility, and (iii) media cloud platform.  Resources at the ROSE Lab consist of NTU & PKU Faculty, Researchers, PhD Students, and Visiting Scholars & Students from other institutions from around the world. 


The successful applicant will be responsible for the support of GPU-based applications (eg. for machine learning and image processing), GPU & cloud systems implementation, and systems management at the ROSE Lab.  This includes:

(i) Hardware installation of servers, GPUs, high-speed networks (Infiniband),
(ii) Storage system implementation and management.
(iii) Systems virtualization (eg. VMware, KVM), and OS (Windows, Linux) installation & configuration.
(iv) Network implementation & management (Infiniband, Cisco switches)
(v) Installation and configuration of machine learning frameworks (eg. cudaConvNet, Caffe) and cloud middleware (eg. Mesos, YARN).
(vi) Management and Monitoring of ROSE Lab’s computing infrastructure
(vii) Provide systems and trouble-shooting support to the ROSE Lab’s IaaS End-users working on machine learning and image processing applications.

Job Requirement:

•  Master degree in Computer Science & Engineering.
•  Good knowledge of x86 computer architecture, server systems architecture, and computer networking.
•  Excellent knowledge of Linux CLI environment, and good experience with CentOS & Ubuntu.
•  Good experience with configuring and managing Network Switches (e.g., Cisco iOS system, Infiniband).
•  Experience with Virtualization platforms (VMware, KVM) and Performance Monitoring tools (Ganglia, Nagios, Libvirt) is preferred.
•  Experience with Machine Learning (eg. Caffe, cudaConvNet) and GPU Computing (CUDA programming) is also preferred.
•  Familiarity with building private cloud with Cloudstack/Openstack + YARN/mesos, and programming experience with Hadoop/Spark would be advantageous.
•  Industry experience would also be advantageous.
•  Strong motivation and initiative of carrying out implementation and research tasks independently.
•  Good interpersonal skills, with the ability to work with people from varied backgrounds.

Learn more about ROSE Lab at

Application procedure:

Interested applicants please attach your full CV, with the names and contacts (including email addresses) of 3 character referees, and all relevant academic certificates to

We regret that only shortlisted candidates will be notified.

Application closes when the position is filled.

Share Article