We are FAST Computing

We develop innovative mathematical models to lead
real innovation in industry

Time reduction: 100 times faster

Energy reduction: 80% less

Everything on a Desktop pc or a laptop

Simulation of structural parametrisation:

Thanks to our expertise we were able to cut time and energy to simulate a complex model like a cruise ship

Image recognition:

A Convolutional neural network to recognise and detect objects in a specific environment: thanks to our expertise we were able to cut time and energy to create and train a complex model

Time reduction: Over 70%

Energy reduction: Over 80%

Parameters reduction: Over 90%

Live Demo: Simulation Speed Comparison

In the comparison slider below, you can see a glimpse of the work we do FAST Computing. The original model takes a full 5 minutes to simulate just one second, while our FAST Computing Reduced Order Model (ROM) accomplishes the same task in 0.1 second of CPU time. This remarkable 50x increase in simulation speed represents a significant advancement, with the potential to improve efficiency and save valuable time and resources across various applications.

How it works

We have developed many libraries based on ROMs. This technology aims to provide solutions with low computational complexity, which means less energy and less time, but always accurate and reliable.

Reduced Order Modeling

We integrate our techniques with Artificial Intelligence and Machine Learning, two powerful tools to improve results.

Machine Learning

We explore several data-driven algorithms to improve the results.

Data Science

We integrate the potentials of high-perfomance computing to develop the most complex projects.

High Performance Computing

We combine Big Data, HPC, ROMs, AI, IoT to build Digital twins, for processes and products, able to prevent damages and guarantee predictive maintenance.

Digital Twin

We exploit the potentials of integration between software and data, provided by sensors installed on devices which are connected to networks.

Internet of Things

UQ is the science of quantitative characterization and estimation of uncertainties in computational applications.

Uncertainty Quantification

Mathematical modeling is a process in which real-life situations and relations in these situations are expressed by using mathematics.

Mathematical Modeling

Some of our projects within SISSA mathLab

Advanced Simulation of Complex Systems

Structural Optimization

Development of a Digital Twin

Web Computational Platforms

We are developing two platforms to let you try the potentials
of reduced order modeling. Few easy prototype tutorials are available here.

ARGOS

ARGOS is the online platform for realtime scientific computing for industrial applications.

ATLAS

ATLAS is the online platform for realtime scientific computing in biomedical applications.

Meet the Team

FAST Computing is a SISSA - Scuola Internazionale Superiore di Studi Avanzati - Startup born within SISSA mathLab, an applied mathematics Lab devoted to develop computational technologies. In the last years, thanks to a grant of 1.6M€ by the European Research Council, we developed AROMA-CFD project, creating many software libraries based on reduced order modelling, which represents our solution for real time simulation. In 2022 we won an ERC PoC award to develop ARGOS, a new web-platform which supplies real-time solutions available on every kind of device.

Give a look at SISSA mathLab website to see our background