Case Study: Erasmus University Rotterdam achieves fast access to processing power for recommendation models with SURF

Easy access to processing power for computational recommendation models

Erasmus University Rotterdam (EUR) researcher Flavius Frasincar and his team needed to improve financial news recommendation systems by using complex computational models that analyze word meanings and relationships. The immense processing power required for these computations made the project infeasible to run on a standard PC, creating a significant challenge for their research. The team turned to vendor SURF for a solution.

SURF provided the university with easy access to its high-performance compute cluster. This service, supported by SURF's technical expertise and training, allowed the researchers to efficiently run their models. The solution accelerated their computational work, enabling the research to proceed after previous attempts had failed, and provided the flexibility to quickly acquire additional computing hours as needed.

Erasmus University Rotterdam

Flavius Frasincar

Lecturer and Researcher