Our large-scale computer systems including vector-typed and scaler-typed computational resources are available for project use in accordance with the purposes of JHPCN and HPCI. To promote research projects which realize excellent science utilizing our supercomputer systems, we call for proposals of research projects using our computational resources. Calls for the research projects, eligibility, and the procedures are described below.
Please see here about projects that have been accepted.
Accepted projects
Exploratory Research Project for Young or Women Scientists
Project representatives | Project |
---|---|
Hajime Shimakawa (School of Engineering, The University of Tokyo) |
Deep-Learning-Based Discovery of Super-Composite Electrical Insulating Materials * [doi.org/10.18910/92759] |
Koichi Mayumi (The institute for sqlid state physics, The University of Tokyo) |
Sliding Dynamics of Ring on Polymer in Polyrotaxane * [doi.org/10.18910/92760] |
Kie Okabayashi (Graduate School of Engineering, Osaka University) |
Development of data-driven cavitation model using CFD database of cavitating turbulent flow * [doi.org/10.18910/92761] |
Hiroyuki Kai (Tokyo University of Science) |
Systematic understanding of the structure-function relationship of droplet transport and collection driven by surface wettability patterning in fractal open microfluidic channels * [doi.org/10.18910/92762] |
* the project gets the acceptance of JHPCN exploratory research project
AI Research projects
Project representatives | Project |
---|---|
Junya Sato (Graduate School of Medicine, Osaka University) |
Deep Learning based radiomics analysis using lung CT datasets and radiology reports. [doi.org/10.18910/92763] |
Harry Handoko Halim (Graduate School of Engineering, Osaka University) |
The Elucidation of Non-equilibrium States of Heterogenous Catalysis by Data-driven Multiscale Simulation: A Case Study of Methanol Synthesis [doi.org/10.18910/92764] |
Takaharu Yaguchi (Graduate School of System Informatics, Kobe University) |
Development of a framework for physical modeling and simulation based on deep learning [doi.org/10.18910/92765] |
Yuji Hamamoto (Graduate School of Engineering, Osaka University) |
Structure search of silicene on the Ag(111) surface based on evolutionary algorithms and Gaussian process regression [doi.org/10.18910/92766] |
The reporting meeting for the research proposal-based use (FY 2022)
Call for research project proposals that require our supercomputer system (FY 2022 add)
1. Overview and Eligibility
The Cybermedia Center basically seeks for proposals from two categories: (1) Exploratory Research Project for Young scientists and Female Scientists; (2) Large-Scale High-Performance Computing Project. Also, the category of (3) AI Research projects is specially set up in FY2022 from our Cybermedia Center's aspect of assisting and contributing AI-related research requiring supercomputing. For example, under this category, we are seeking for material scientific research using deep learning for material design, bioinformatics research using machine learning for compound search, humanities and social science research necessitating large-scale statistical processing, computer engineering utilizing computation/data-intensive data analysis methods and so forth. We are calling for research proposals using AI technologies from every field of science.Details of eligibility are described below. Only the applicants who have applied to either JHPCN(except for a JHPCN exploratory research project) or HPCI are permitted to submit to the FY 2022 research proposal-based use, in the case that the applicants have used our HPC systems through FY2021 research proposal-based use.
(1) Exploratory Research Project for Young Scientists and Female Scientists
- The principal investigator of the project proposals should be a researcher age 42 years old or younger as of April 2021, or a female researcher.
- The applicant must agree to the following conditions. If an applicant's proposal for a research project meets the conditions of JHPCN and if their proposal is accepted for a JHPCN exploratory research project, the proposal may be also accepted as a JHPCN Exploratory Research Project.
(2) Large-Scale High-Performance Computing Projects
- Proposals must include a plan for enhancing the degree of parallelism of a highly parallel execution application/simulation solved utilizing our large-scale computer systems.
(3) [special] AI Research projects
- Proposals must include a plan for enhancing the size of research problem by executing AI-related programs on our large-scale computer systems.
2.Acceptance numbers and allocated resources
Scheduled numbers to be adopted
(1) Exploratory Research Project for Young Scientists and Female Scientists | a small number |
(2) Large-Scale High-Performance Computing Project | |
(3) [special] AI Research projects |
Approximate resource upper limit providing for each category
- † Not the upper limit of an application
resource name | (1) | (2) | (3) | |
---|---|---|---|---|
SQUID | General purpose CPU nodes | 1,000,000 node-hours |
1,000,000 node-hours |
500,000 node-hours |
GPU nodes | 50,000 node-hours |
50,000 node-hours |
25,000 node-hours |
|
Vector nodes | 35,000 node-hours |
35,000 node-hours |
20,000 node-hours |
|
OCTOPUS | General purpose CPU nodes | 210,000 node-hours |
210,000 node-hours |
100,000 node-hours |
GPU nodes | 30,000 node-hours |
30,000 node-hours |
20,000 node-hours |
|
Xeon Phi nodes | 40,000 node-hours |
40,000 node-hours |
20,000 node-hours |
|
Large-scale shared-memory nodes | 2,000 node-hours |
2,000 node-hours |
1,000 node-hours |
resource name | (1) | (2) | (3) |
---|---|---|---|
storage for SQUID | 1,000TB | 2,000TB | 1,000TB |
storage for OCTOPUS | 100TB | 200TB | 100TB |
3. Project Implementation Period
Mid-June, 2022 - March 31, 2023
If you require utilizing the computational resources over the period, you should apply a continuance application, then you should get an acceptance.
4. System Overview
Two types of large-scale computer systems are available.
"SQUID" is the newest system in our center that composed three types of nodes, general purpose CPU nodes, GPU nodes and Xeon Phi nodes.
"OCTOPUS" is a scalar computer cluster that composed four types of nodes, general purpose CPU nodes, GPU nodes, Xeon Phi nodes and large-scale shared-memory nodes.
5. Range of Support Services
In this application, the range of support services are described below.
- The large-scale computer system usage fee for the proposed research project is supported by the Cybermedia Center.
- The meeting room in the Cybermedia Center can be used for the proposed research project.
- The travel bills for oral/poster presentation of the proposed research project is supported by our Center.
- Tuning support is supported in response to the applicant’s needs.
6. Proposal Evaluation
The high performance computer system committee of our Center will review and evaluate the received proposals. The standards for review and evaluation are described below.
(1) Exploratory Research Project for Young Scientists and Female Scientists
- The proposed research project utilizing our large-scale computer systems holds likely promise for the future.
- The scale of the problem can only be solved with our large-scale computer systems.
- The schedule for use of our large-scale computer systems is feasible.
- The proposed research project has a continuity of supercomputer use.
(2) Large-Scale High-Performance Computing Project
- The proposed research project through the use of our large-scale computer systems hold the promise of making an important impact on science.
- The program used in the proposal has a high-scalability of parallelization.
- The schedule for use of our large-scale computer systems is feasible.
(3) [special] AI Research projects
- The proposed research project through the combinational use of AI technologies and our large-scale computers hold the promise of making an important impact on science.
- The scale of the problem can only be solved with our large-scale computer systems.
- The schedule for use of our large-scale computer systems is feasible.
7. Responsibilities of Selected Applicants
- Selected applicants must present their research achievements in the reporting meeting (scheduled in March 2023).
- Selected applicants must write a scientific article about their research work for the "HPC Journal" published in the Cybermedia Center.
- Selected applicant must use our template format for all research reports.
- Selected applicant must make a presentation in the JHPCN symposium if the project is accepted as a JHPCN’s Exploratory Research Project.
- Selected applicant must apply for JHPCN or HPCI assignment for the next fiscal year. (Only for applicants accepted in (1) or (2))
- When publishing research achievements that have been obtained as a result from our computer resources, the selected applicant must indicate that his/her research was supported. A sample acknowledgment follows:
-
Sample Acknowledgement
This work was (partly) achieved through the use of Supercomputer System SQUID at the Cybermedia Center, Osaka University.
If you do not make a presentation at the reporting meeting or contribution to the HPC journal, you will not be able to apply for the next year. Additionaly, participation as a collaborator is also not possible.
8. Contract Stipulation
The usage rules of our supercomputer systems are described below.
Rules
9. Application Method and Period
Please send an e-mail with your application from to the address below. Please use the following application form.
Application Form Application Form (AI Research projects)
Address | Information and Communications Technology Services Infrastructure Division, Cybermedia Center, Osaka University |
system@cmc.osaka-u.ac.jp | |
Opening Date | March 14, 2022 |
Closing Date | April 15, 2022 |
10. Notification
Receipt of a notification of acceptance of the researcher's project proposal also includes an announcement of computer resource use.
Notification Date | Early-June, 2022 |
11. Contact Information
If you have questions regarding this application, please contact us at the email or address below:
Consultation/Inquiry | E-mail
Address
Postal Code
|
FAQ
Q.I was adopted in resrch proposal based use this year and I use supercomputer now. Can I apply resrch proposal based use of next year in a row?
A. You can apply on the condition that you have already applied to HPCI or JHPCN of next year. Please note that it is not possible to apply the same project in the same fiscal year in additional accepting.
Q.In 7. Responsibilities of Selected Applicants, it is mandatory to apply for JHPCN or HPCI assignment for the next fiscal year. Does an applicant of the Research Proposal-based Use need to apply as a “representative” for JHPCN or HPCI?
A. No, that is not necessary. Participation as a collaborator is no problem.
Q.I'm adopted in not only research proposal based use, but also HPCI or JHPCN in the same fiscal year. Can I use both together?
A. Yes. However, they expect adopted users to use up allocated resources. If you are uneasy about whether or not to use up, you can decline the reserch proposal based use.
Q.Can a graduate student apply?
A. Yes, but a graduate student is required to enter the information of a supervisor when apply.