Guidelines
Data deletion and extension policies
Every year at 1. August, all student data and user information will be automatically removed from the AI-LAB platform. You will receive email notifications 2 months, 1 month, 14 days, and 2 day prior to the deletion date. It is important to ensure that you transfer any desired files from AI-LAB to your local computer before the deletion date.
Extension request:
Should you wish to keep your data on AI-LAB and use the platform for another year, you can submit an extension request through the application form for AI-LAB. This will extend your access until August 1st of the following year.
Not for confidential or sensitive data
In AI-LAB you are only allowed to work with public or internal information according to AAU’s data classification model (classified as levels 0 and 1, respectively).
If you would like to work with confidential or sensitive data (classified as levels 2 and 3), then we support another HPC platform called UCloud.
Scheduled maintenance
On specific dates, scheduled maintenance will be conducted on AI-LAB. The routine maintenance will occur between 00:01 and 23:59. AI-LAB will be unavailable throughout most of that day. You can still submit new jobs until the beginning of the service window. For jobs that may exceed the service window, please ensure to set a maximum runtime using the parameter --time that concludes before 23:59 the day preceding the service window. Read more about the --time parameter here. Otherwise, these jobs will not be able to start until after the maintenance period. You will receive email notifications 1 month, 14 days, and 1 day prior to the scheduled maintenance window.
Clean up after use
AI-LAB encourages responsible usage, which includes cleaning up your user directory for files that are no longer needed. You can use rm [file_name] to remove a file and rm -r [directory_name] to remove a directory and its contents.
Before deleting files, remember that you can transfer files from your AI-LAB directory to your local computer.
Data Management best practices
Effective data management is crucial for maximizing the value of your research and ensuring its integrity. Organize your data in a clear and structured manner, labeling files and directories appropriately. Regularly back up your data to prevent loss in case of unexpected events. Additionally, adhere to any data sharing and privacy regulations applicable to your field. Get more guidance about data management here.
Jobs can run no longer than 12 hours
Every job submitted to AI-LAB is subject to a time limit of 12 hours. This limit is set to prevent a single user from monopolizing the entire cluster indefinitely. We are trying to ensure that all users receive an equal share of available resources.
Not intended for CPU processing
The platform excels in tasks requiring parallel processing capabilities, such as training deep learning models, image and video analysis, and other GPU-accelerated computational work. However, it's not intended for applications that only need CPU processing.
Allocating the right amount of resources
Please be mindful of your allocations and refrain from allocating many resources without knowing/testing/verifying that you indeed can utilise all of the allocated resources.
Please be mindful and de-allocate the resources when you are no longer using them. This enables other users to run their jobs.
Focus on data processing, not storage
AI-LAB is designed primarily for data processing, not storage. Our High-Performance Computing (HPC) platform supports complex data analysis and simulation tasks efficiently. This focus ensures our resources are distributed optimally among users.