Welcome to the Open Science Cloud research laboratory

The Open Cloud Research Lab is the largest open research e-Infrastructure in the state. The laboratory was procured with the support of the Ministry of Education and Science, through the competition for financing scientific research projects of special and public interest for 2021 (support for the development of laboratory resources) (70%), and co-financing by FINKI at UKIM (30%).

The laboratory consists of several components:

  • GPGPU cluster
  • Openstack cloud
  • Data storage

The first unit represents a high-performance computing cluster based on graphic processors, which in the past period have become the primary means of processing big data and artificial intelligence (Big data, Artificial Intelligence). The second unit is a cloud computing platform that will provide fast and simple access to computing resources for researchers. The third unit will represent a large data warehouse, which is integrated with the previous two units, and in which researchers have the opportunity to store their data sets. Also within the data storage space, a repository of open data sets (open datasets) has been established where researchers can publicly and openly publish data sets and results, which will increase their visibility and the possibility of collaboration and implementation the postulates of open science.

The Open Science Lab consists of two main systems, each with a specific purpose. The first system is an open cloud aimed at country researchers, enabling the deployment of their own virtual infrastructure for high-performance computing. All work with the system is based on the self-service strategy, where users independently manage their personal resources, from setting up virtual machines, through network configuration, to adding additional storage space. The modern computer equipment on which this system is installed guarantees high performance, with a total of 640 processor cores and over 15 TB of working memory.

The second system is intended for performing complex computational tasks, with the additional possibility of using graphic processors. Researchers can independently schedule the execution of a complex task, choosing whether or not they need dedicated GPUs. The system is equipped with state-of-the-art hardware, including a total of 8 GPUs with a total video memory of 640 GB, making it suitable for application in researching problems from a variety of fields, where the application of GPUs would speed up calculations.

In order to enable smooth operation with big data, both systems have a fast connection to a centralized storage space that offers over 2 PB of total space available. The use of this space can be done in a number of ways, depending on the needs of the researchers, including for: running virtual machines, long-term storage of data sets, storage of input and output data during their analysis, and general access through a programming interface (API) for integration with third-party computing systems.