CCMS

Center for Computational Materials Science
Institute for Materials Research,
Tohoku University

Supercomputing system trial usage

1. Outline

"Can I run a program that I used on my PC with the supercomputer?"
"Can applications available at the Center for Computational Materials Science be used for my research?"
There is a demand for trying these and others.

With this usage frame, you can try out the supercomputing system for up to two weeks. Currently, it takes time to apply for the supercomputing system because it is necessary to apply for the research project, to examination and to adopt it.
Therefore, we will provide "Supercomputing System Trial ". Applicants can use the supercomputing system for up to two weeks before applying for joint use issues. We hope to improve convenience for researchers and promote research.

2. Qualification for use

Those who fall under any of the following and do research on materials science field.

  1. Those who belong to Institute for Materials Science, Tohoku University
  2. Researchers belonging to national, public, and private universities and national and public research institutes

3. Available computing resources

Supercomputing system of Center for Computational Materials Science, Institute for Materials Research, Tohoku University (HPE Cray XD220v, HPE Scale-up Server 3200, and HPE Cray XD670)

System Name Supercomputer
Model HPE Cray XD220v HPE Scale-up Server 3200 HPE Cray XD670
total nodes

General node : 120
Large memory node : 7

2 nodes

11 nodes
CPU

Intel Xeon Platinum 8480+

  1. Frequency:
     2.0 GHz
  2. Number of cores:
     56 cores
  3. Number of CPUs:
     2 CPUs/node

Intel Xeon Platinum 8490H

  1. Frequency:
     1.9 GHz
  2. Number of cores:
     60 cores
  3. Number of CPUs:
     16 CPUs/node

Intel Xeon Platinum 8480+

  1. Frequency:
     2.0 GHz
  2. Number of cores:
     56 cores
  3. Number of CPUs:
     2 CPUs/node
Accelerator - -

NVIDIA H100 80GB SXM5

  1. Performance:
     33.5 TFLOPS(FP64)
  2. Number of GPUs:
     8 GPUs/node
  3. Memory bandwidth:
     3.352 GB/s
memory size 512 GiB/node
2.0 TiB/node		 4.0 TiB/node 1.0 TiB/node
Theoretical peak performance 4.052 PFLOPS
(CPU : 1.104 PFLOPS, GPU : 2.948 PFLOPS)

4. Node hour and usage fee

4.1 Available node hour

*The supercomputer's node time is up to 50 hours.
When using the application for PC, use the computer installed in Supercomputer building development room 1.
Trial availability time: from 9:00 to 17:00 except Saturdays, Sundays, and holidays

4.2 Usage fee

Usage fee would not be collected.

5. Rules of trial usage

  1. The maximum trial period is two weeks
  2. Applications are limited to once per year, so applicants can only try
  3. As for the result obtained in the trial period, it can not be announced at the paper, the conference etc. If it is to be announced, please apply for collaborative research at the CCMS
  4. You can not copy, modify, redistribute applications
  5. No report submission is required for the work done during the trial period
  6. We will not assume any responsibility for any loss or damage caused by trial of this system
  7. If you wish to use the supercomputing system after the expiration of the expiration date, you will need to apply for collaborative research on CCMS, IMR
  8. Travel expenses etc are not paid to users

6. How to apply

Click on the "Application for trial" for the flow on the previous page.