filter >

Photo

Cheap and Fast Mathematical Optimisation with Cloud Functions

Emanuele Fabbiani

from xtream (Italy)

About speaker

Head of AI at xtream

Engineer, researcher, entrepreneur. Emanuele earned his PhD in AI by researching time series forecasting in the energy field. He was a guest researcher at EPFL Lausanne, and he's now the Head of AI at xtream, where he solves business problems with AI.

About speakers company

xtream is a small software house in Italy that develops AI solutions. Founded in 2018, it is now a partner of corporations such as Plenitude, A2A, Credem, Yamaha, Lavazza, Illimity, and scaleups such as WeRoad, Credimi, and Senso.

Abstracts

broad

This talk showcases the use of serverless computing to solve large-scale optimization problems efficiently.

Starting from a real case study, we discuss different cloud-based architectures and we show how cloud functions enable a 96% reduction in elapsed time while keeping the cost low.


Solving large-scale optimization is challenging.

However, some optimization problems can be easily parallelized. In this case, serverless computing can drastically cut the elapsed time, while keeping the cost low.

In this talk, we consider a real case study. While creating power market scenarios, we need to find the optimal way of using power storage facilities.

This results in a convex optimization problem, where the power exchanged between each facility and the grid in each hour is an optimization variable. Over a scenario range of 50 years, this generates hundreds of thousands of variables.

Fortunately, the problem can be split into subproblems which can be solved in parallel.

Nonetheless, a conventional solution, based on a 16-core virtual machine, requires about 2 hours to perform the computations. We show how a serverless approach based on cloud functions is capable of cutting the elapsed time by 96% while keeping the operating costs as low as 3€ per run.

Finally, we show how to connect the parallel optimization module to the rest of the scenario processing pipeline, implemented in AirFlow.

After attending this talk, you will be aware of the capabilities of serverless for high-performance computing, and you will have a better feeling about how to design a cloud architecture.

The talk was declined

other talks of this topic

Photo
Architectures that we can use with .NET

Alexej Sommer

Capgemini

broad
Photo
ML/AI in the cloud - State of the Art in 2025

Federico Fregosi

OpsGuru

broad
Photo
Beyond Caching: Valkey's Advanced Data Structures in Action

Viktor Vedmich

Amazon Web Services

specific
Photo
The Anatomy of a Distributed JavaScript Runtime

Peter van Vliet

Masking Technology

broad
Photo
Cloud Costs with ClickHouse and OpenCost

Denys Kondratenko

Altinity

specific
Photo
REST or gRPC: Best practices for modern architectures

Kristina Kraljić

PIS d.o.o.

specific
Photo
Mastering Software Design: Best Practices for Building Robust Applications

Ambesh Singh

Visionet Systems Deutschland

broad
Photo
DRBD Deep Dive

Philipp Reisner

LINBIT

specific
Photo
Open Source Ecosystem for ClickHouse on Kubernetes

Denys Kondratenko

Altinity

specific
Photo
Achieving True Layered Separation with Hexagonal Architecture in Spring Boot

Adrian Kodja

softgarden e-recruiting GmbH

specific
Photo
Organizational Sustainability with Platform Engineering

Lesley Cordero

The New York Times

specific