NUS IT’s Research Computing team provides High-Performance Computing and Artificial Intelligence/Machine Learning (AI/ML) resources and services to researchers through on-campus infrastructure or by partnering with commercial cloud service providers. One of our missions is to promote advanced computing technologies to accelerate researchers’ works, enabling them to gain a competitive edge in their research fields. 

This program aims to identify innovative and impactful projects within the NUS research community.  Eligible projects that are successful in this call for applications will be awarded one of the following privileges:

• Priority H100 GPU Access – Priority use of our latest Nvidia H100 GPU resources for 6 months.
• Cloud Credits for Research Program – Free, one-off AWS cloud credits to help kick-start the project in the cloud.

Priority H100 GPU Access

Successful projects will be given access to our cluster of six Nvidia H100 servers (total 48 GPUs) suitable for running AI/ML workloads.  This is on top of other HPC resources which is available to all HPC users.  The expected launch date of our new GPU servers is in early April 2024.

Cloud Credits for Research Program

The cloud credits can be used by researchers to consume a wide range of AWS cloud resources and services (some limitations may apply). This program is modelled after the AWS Global Cloud Credit for Research Programme and is customised specifically for NUS researchers.  Last year (2022), a total of USD200,000 of cloud credits were awarded to 19 projects in this program. 

• This call for applications is open to all NUS researchers (staff or students).  Projects under the Cloud Credits for Research Program have to be proposed with AWS accounts that are registered using NUS email addresses, subject to validation by the evaluation committee. If the researchers do not have AWS accounts yet, the research computing team can assist to create them post award.
• Applicants shall be either research staff or PhD students. PhD students will need to provide proof of support from their supervisors / PIs.
• The Principal Investigator (PI) must have a full-time academic appointment at an NUS Faculty / School (Assistant Professor, Associate Professor or Professor) or Senior Research Fellow or above in Research Institutes / Centres in NUS.

• Successful projects will be given early access to our new H100 GPU cluster, with the following allocated benefits:
  • 6 months of priority GPU usage
  • Up to 8 Nvidia H100 GPUs per project
  • Up to 12 CPU-cores and 256GB RAM per GPU per project
  • Up to 10,000 GPU-hours per project
  • 2TB of workspace per project
• Successful applicants will receive training and support for running their workloads in the new H100 GPU cluster.
• At the end of the 6-month priority usage period – GPU systems will be open to all HPC users, and all jobs will be subjected to fairshare allocations and policies.

• Each successful application under the Cloud Credits for Research Program will be awarded up to a maximum of USD5,000 credits, or an amount deemed suitable by the evaluation committee.
• For applications requiring cloud credits exceeding USD5,000, further evaluation will be done on a case-by-case basis.

Submitting an Application

Priority H100 GPU Access

Proposers shall use the online form to submit their project applications:

1. Each application shall include the following information:
   1. Description of the problem to be solved.
   2. Proposed solution.
   3. Proposed use of GPU resources (including specific software tools to be used) – basic toolchain will be made available.
   4. Project deliverables.
   5. Potential benefits and future use of project outcomes.
   6. Plan for sharing outcomes (tools, data, and/or resources) created during the project.
2. Submit an estimate of the project duration and key milestones.

Cloud Credits for Research Program

Applicants shall use the online form to submit their project applications.

1. Each application shall include the following information:
   1. Description of the problem to be solved.
   2. Proposed AWS solution (including specific AWS tools, timeline, key milestones).
   3. Plan for sharing outcomes (tools, data, and/or resources) created during the project.
   4. Any potential use of AWS beyond supported project duration by individual research group or broader community.
   5. Any AWS Public Datasets to be used in your research.
2. Submit an estimate of the AWS resource usage via the AWS Pricing Calculator and submit the link to the online calculations.

Submission Timelines

1. This Call for Application is open until 2359 (SG time), 20 December 2023.
2. Applicants may be invited for an interview with our evaluation committee to answer questions about their proposed projects.
3. Applicants will be informed of the evaluation committee’s decision in early 2024.

Call-for-Proposal Briefing

1. A briefing for the HPC Priority Projects for Research Program will be held on 4th December 2024, 3:30pm in the NUS IT Orange Carrot Meadow, Level 1, NUS IT Building.
2. The program is currently closed.  An online form will be provided when the briefing is planned.

Project Completion Report

Successful awardees will be required to submit a report and may be requested to present their work upon completion of their project, and to acknowledge NUS IT in their publications using grant numbers:

• NUSREC-HPC-00001 (for Priority GPU Program)
• NUSREC-CLD-00001 (for Cloud Credits Program).

For any clarifications on the above, please email: nusit-rec-support@nus.edu.sg