Skip to main content

NIG Supercomputer (2019 => 2025)

The National Institute of Genetics (NIG) provides a cutting-edge supercomputer system equipped with large-scale cluster-based computation, large-scale shared memory computation, and high-capacity high-speed disk devices, serving as a large-scale computational infrastructure for information processing in life sciences and medical research.

top_image

Important

Announcement: Replacement of the Current NIG Supercomputer

The current NIG supercomputer, which began operations on March 1, 2019, will be replaced by the next-generation NIG supercomputer on March 1, 2025.

1. Configuration of the Next-Generation Supercomputer

The next-generation supercomputer will retain the basic configuration of the current system while enhancing the network connection bandwidth with SINET6 to 100Gbps. The computational performance per CPU core will be approximately doubled.

For GPU-equipped nodes, the system will introduce NVIDIA DGX B200 for cutting-edge AI analysis, as well as NVIDIA L40S and PEZY-SC3 nodes for cost-effective genome analysis.

For detailed configuration of the next-generation supercomputer, please refer to the Hardware (2025) page.

2. Schedule for Activation

  1. Data migration of users' home directories will begin in late December 2024. The data in the home directories as of February 14, 2025, 5:00 PM will be copied to the new supercomputer's disk. From 14th February 2025 onwards, please ensure that you back up your data as required and transfer it to the new supercomputer at your earliest convenience.
  2. The current supercomputer will be shut down at 9:00 AM on February 21, 2025, to switch to the next-generation supercomputer. Login services for the new supercomputer will commence at 12:00 PM on March 1, 2025. The current NIG supercomputer will be unavailable from 9:00 AM on February 21, 2025 until midday on March 1, 2025.
  3. GPU computation nodes with L40S will begin service on April 1. During the period from March 1 to March 31, GPU computation nodes will not be available.
  4. GPU computation nodes with B200 are scheduled to begin service on June 1.

3. Important Notes on Migration to the Next-Generation Supercomputer

(1) The job scheduler will change from Grid Engine to Slurm. Job scripts for Grid Engine need to be rewritten for Slurm. Please refer to the Software > Slurm page for details on transitioning from Grid Engine to Slurm.

(2) The operating system of all compute nodes in the next-generation supercomputer will be updated from the current Ubuntu Linux 22.04 to Ubuntu Linux 24.04.

(3) The pricing for storage, memory-optimized nodes, and accelerator-optimized nodes (GPU nodes) will be revised based on a proportional lease cost approach for business continuity. The revised pricing details will be published on the homepage soon.

warning

Due to limited disk capacity, the research institute does not back up data in users' home directories. Users are responsible for backing up their data.

Year-End Account Renewal Applications

The account application system is currently being revised for security enhancements, which is causing a delay in its reopening. Account renewal applications will be processed after the system is reopened. For this fiscal year, account renewal for the next-generation supercomputer will be handled by the supercomputer administrator without use's application.

Regular Maintenance in Fiscal Year 2024

Regular maintenance for fiscal year 2024 was conducted from December 13 to December 19, 2024.

Temporary Suspension of New Account Registrations for Large-Scale Users in General and Personal Genome Analysis Areas (May 13, 2022)

Thank you for using the NIG supercomputer.

Due to the supercomputer's computational resources being highly constrained in terms of both CPU and disk, new account registrations for the following users will be temporarily suspended:

  • Large-scale users in the general analysis area
  • Users in the personal genome analysis area

The following applications will continue to be accepted:

  • Additional resource usage requests for users who are already large-scale users in the general analysis area or personal genome analysis area.
  • Regular user registrations for the general analysis area. In this case, disk usage is limited to an initial setting of up to 1TB.
  • Applications for workshop usage will also continue to be accepted.

However, we ask for your understanding as it may not always be possible to meet your requests.

The resumption of new account registrations will be decided based on the availability of computational resources. Updates on the availability will be published on the homepage every four months. Prospective users are requested to refer to these updates and submit their usage plans in advance. Please note that the availability forecasts are not guaranteed, and requests may not always be fulfilled.

Acknowledgements

The activities of the NIG supercomputer system are evaluated through the acknowledgements received. If a paper utilizing this supercomputer system is accepted, please include an acknowledgement using the examples below.

Examples:

==============================================

<English Acknowledgement>
Computations were partially performed on the NIG supercomputer
at ROIS National Institute of Genetics.

==============================================

<Japanese Acknowledgement>
本研究は、情報・システム研究機構 国立遺伝学研究所が
有する遺伝研スーパーコンピュータシステムを利用しました。

================================================