ETAIS is a project providing R&D groups with computational and storage resources to do innovative and useful (or just innovative) projects. We have listed several use cases that could explain better why you would need ETAIS. If some of them feel appealing please apply for ETAIS now!
Modelling of experiments in chemical, genome, material science and many more branches of science requires extensive computational resources, both CPU and memory. ETAIS is offering a heterogeneous infrastructure that optimally matches requirements of a specific modelling task.
Examining multiple scenarios with changing parameters is a technique often used for finding an optimal solution to a problem, e.g. finding an optimal circuit design or the most efficient construction plan. ETAIS is providing batch processing resources that allow to enqueue combinations as separate jobs and process them in a parallel over 15000 cores.
ETAIS offers a wide variety of pre-installed packages for statistical analysis. Low-latency network connection allows to run MPI-based jobs or clustered analysis (R, Jupyter) efficiently and in real time.
Optimization and predictions
ETAIS large-memory nodes and GPU infrastructure allows to train and process larger deep learning models that are typical in linguistics, biology, computer vision and many other areas of science.
Processing and preservation of Big Data
ETAIS provides a tiered shared storage that allows to preserve and process large amounts of data (in Petabytes).
Internet of Things
ETAIS is connected to several Internet networks, as well as campus network, which makes it a great platform for IoT aggregators and analysis solutions.
Machine learning and computer vision
ETAIS is investing a lot in expansion of its GPU cluster as such infrastructure is increasingly important for a number of applications in machine learning and computer vision, allowing to significantly reduce training time for neural networks as well as provide services with very low latency (e.g. processing of webcam streams, object and landmark recognition, online translations).
ETAIS infrastructure can be used both for modelling of the risk events as well as performing vulnerability testing, e.g. estimation of organisation’s password strength using GPU-accelerated password crackers.
3D Animation rendering requires a lot of memory and computational resources that ETAIS is happy to provide, allowing to create Full HD animations within reasonable time and budget.
Long term data preservation
ETAIS is providing a tape library that allows to put data to an offline storage (potentially also moving it into a remote location) and keep the archive for the years to come.