The BDSRC provides high-end computing resources, high-capacity storage, and secure instant backup. BDSRC centrally manages resources to ensure the reliability and security of the compute and storage systems. Researchers can remotely access our computing resources at any time through network connections.
Get an Account
First, a principal investigator needs to fiil the application to request access to the Computing System of Research for his/her research team or project. The application form includes the research purpose, usage plan on the system, list of team member, etc. Applicants should send the completed form to ntubdsrc@ntu.edu.tw, and we would reply the review result in 10 days.
Second, a member from an approved team needs to submit the online form to request a personal account. Please allow three business days for a response.
Every applicant should check the latest regulation on Computing for Research which announced on Jan 20th, 2025.
Get Support
Please follow the User Guide to access to the computing system.
If you encounter any problems, user support is available by email at ntubdsrc@ntu.edu.tw or phone at 02-3366-8300 #55808.
Hardware Configuration
Our infrastructure is built to provide exceptional performance for AI, data analytics, parallel computing, and other compute-intensive tasks, surpassing the capabilities of standard PCs and laptops.
- Powerful Compute Resources: Equipped with high-core-count CPUs and substantial memory, enabling efficient handling of complex, resource-heavy workloads.
- Cutting-Edge GPU Support: Multiple GPU-optimized nodes, featuring advanced NVIDIA TITAN V and RTX A6000 GPUs, to accelerate AI training and high-performance computing.
- Expansive Data Storage: Over 100TB of storage capacity, allowing seamless processing of large datasets without storage limitations.
- Blazing-Fast Connectivity: High-speed, stable internet across the campus ensures smooth, uninterrupted operations.
master
CPU: Intel(R) Xeon(R) Gold 6242*2
Total cores: 32
Avaliable Memory: 1006GB
Total Memory: 1TB
node01
CPU: Intel Xeon E5-2650 v4*2
Total cores: 24
Avaliable Memory: 252GB
Total Memory: 256GB
GPU: NVIDIA Quadro K620
node02
CPU: Intel Xeon E5-2637 v4*2
Total cores: 8
Avaliable Memory: 252GB
Total Memory: 256GB
node03
CPU: Intel Xeon E5-2650 v4*2
Total cores: 24
Avaliable Memory: 252GB
Total Memory: 256GB
GPU: Nvidia GT218 (GeForce 210)
node04
CPU: Intel(R) Xeon(R) Gold 6326*2
Total cores: 32
Avaliable Memory: 131GB
node05
CPU: Intel(R) Xeon(R) Gold 6326*2
Total cores: 32
Avaliable Memory: 522GB
gpu01
CPU: Intel Xeon E5-2650 v4*2
Total cores: 24
Avaliable Memory: 504GB
Total Memory: 512GB
GPU: NVIDIA TITAN V*2
gpu02
CPU: Intel Xeon E5-2650 v4*2
Total cores: 20
Avaliable Memory: 502GB
Total Memory: 512GB
GPU: NVIDIA 3090*2, RTX A4000
gpu03
CPU: Intel(R) Xeon(R) Gold 6226R*2
Total cores: 32
Avaliable Memory: 754GB
Total Memory: 768GB
GPU: NVIDIA RTX A6000*2
gpu04
CPU: INTEL(R) XEON(R) GOLD 6526Y*2
Total cores: 32
Avaliable Memory: 768G
Total Memory: 791GB
*Application required to access GPU04.
Supported Software
Python 3.10.12, Matlab 2024b, Stata 17.0, R 4.4.1
Acknowledgment
If you have published a paper or have given a presentation that acknowledges the BDSRC's computational resources or other services, please email us with the paper title, journal, or conference where a presentation was given. These citations help BDSRC demonstrate the role of computational resources and support staff in research at NTU. If you would like your BDSRC-supported work featured on our website, please indicate that you are comfortable with us doing so.
Please reference BDSRC as "the Behavioral and Data Science Research Center at the National Taiwan University" in citations and acknowledgments. Some example citations are shown below.
- This research was partly supported through computational resources provided by the Behavioral and Data Science Research Center at the National Taiwan University.
- We are grateful for the support of the Behavioral and Data Science Research Center at the National Taiwan University for assistance with the calculations carried out in this work.
- We acknowledge the Behavioral and Data Science Research Center at the National Taiwan University to support this work.