HEROES · Project

Smart Software That Picks the Cheapest, Fastest Supercomputer for Your Workload

digitalPilotedTRL 6

Imagine you need serious computing power — like simulating how air flows around a car or predicting wind energy output — but you have no idea which supercomputer or cloud service to use. HEROES built a kind of travel-booking engine for supercomputing: you describe your job, and it finds the best platform based on whether you care most about speed, cost, or energy savings. It even handles the messy technical details of submitting your work, so engineers can focus on engineering instead of wrestling with HPC systems.

By the numbers

European SMEs in the consortium building this for their clients

80%

industry ratio in the consortium

partners across 3 countries (Spain, France, Italy)

project deliverables produced

years of development

The business problem

What needed solving

Companies that need supercomputing power — for simulations, AI training, or complex modeling — face a confusing landscape of HPC centers and cloud providers. Choosing the wrong platform means wasted budget, wasted energy, and missed deadlines. Most engineers are not HPC specialists, so they either overpay for cloud resources or underutilize available supercomputing capacity.

The solution

What was built

A decision module prototype integrated into the HEROES platform that automatically selects the best computing platform (HPC center or public cloud) based on user constraints like energy efficiency, price, or performance. The platform also simplifies job submission so users need minimal HPC expertise. A total of 16 deliverables were produced across the 2-year project.

Audience

Who needs this

Renewable energy companies running weather simulations and site assessmentsAutomotive and aerospace manufacturers running CFD and structural simulationsHPC service providers wanting to offer smarter platform brokering to clientsEngineering consultancies that outsource heavy computation to cloud or HPCAI/ML teams at mid-size companies needing cost-efficient access to GPU clusters

Business applications

Who can put this to work

Renewable Energy

SME

Target: Wind or solar energy companies running weather simulations and site assessments

If you are a renewable energy firm spending heavily on weather simulations and site feasibility studies — this project developed a decision module that automatically selects the most energy-efficient or cost-effective HPC platform for your workloads. Instead of your engineers manually comparing cloud providers and supercomputing centers, the software picks the best option and submits jobs with minimal user interaction. The consortium worked with Meteosim and UL Renewables on exactly these kinds of workflows.

Automotive & Manufacturing

mid-size

Target: Vehicle manufacturers and engineering firms running design simulations

If you are a manufacturer using computational fluid dynamics or structural simulations to design more energy-efficient products — HEROES built a platform that routes your simulation jobs to the best-fit computing resource, whether that is an HPC data center or a public cloud. The project specifically addressed manufacturing workflows for designing energy-efficient vehicles, with advisory input from Dallara (racing car engineering). This means faster turnaround and lower compute costs without needing in-house HPC expertise.

IT Services & Cloud Computing

SME

Target: HPC service providers and managed cloud companies

If you are an IT service company that helps clients access supercomputing resources — HEROES applied marketplace concepts to HPC, creating a brokering layer between users and computing platforms. The consortium of 4 SMEs that sell HPC services to clients built this to solve a daily market demand they face. Integrating this decision module into your offering could let you serve more clients with automated platform selection and job optimization.

Frequently asked

Quick answers

What does this actually cost to use?

The project data does not specify licensing fees or subscription pricing. The consortium of 4 SMEs intended to commercialize the outcomes after the project ended in February 2023. Contact the coordinator for current pricing and licensing terms.

Can this handle real industrial-scale workloads?

Yes — the platform was designed specifically for petascale and exascale HPC and ML workloads. The decision module prototype was integrated into the HEROES platform and tested with real workflows from renewable energy (Meteosim, UL Renewables, EDF) and manufacturing (Dallara) advisory partners.

What is the IP situation — can we license this?

The project was funded as an Innovation Action with a consortium of 4 SMEs, suggesting commercial licensing was always the plan. The objective explicitly states that outcomes will be commercialized. Specific IP and licensing terms should be discussed with the coordinator UCIT.

Does it work with any HPC center or only specific ones?

Based on the project objective, HEROES supports submission to both HPC data centers and public cloud infrastructures. The platform is designed to broker across multiple computing platforms rather than lock users into a single provider.

How long does it take to integrate into our existing workflow?

The decision module was designed to require as little user interaction as possible — users describe their constraints (best energy efficiency, best performance-to-price ratio, best price) and the system handles platform selection and job submission parameters. Based on available project data, specific integration timelines are not documented.

Is it still being developed after the project ended?

The project closed in February 2023. The consortium of 4 SMEs had commercialization plans, but current development status should be verified directly with the coordinator. The project website (heroes-project.eu) may have updates.

Consortium

Who built it

This is an unusually industry-heavy consortium: 4 out of 5 partners are companies (80% industry ratio), with 3 being SMEs. There are zero universities — the only non-industry partner is a supercomputing research center providing energy management and resource optimization expertise. All 5 partners come from 3 Southern European countries (France, Spain, Italy). The coordinator UCIT is a French SME. This composition signals a project built to ship a product, not publish papers — these are HPC service companies solving a problem they face every day with their own clients.

NEOVIA INNOVATIONparticipant · FR
BARCELONA SUPERCOMPUTING CENTER CENTRO NACIONAL DE SUPERCOMPUTACIONparticipant · ES

How to reach the team

UCIT is a French SME that coordinated the project. Use the CORDIS contact form or search for UCIT's team via their company website to reach the project lead.

Order a report on this consortium

Next steps

Talk to the team behind this work.

Want to explore whether HEROES technology fits your computing needs? SciTransfer can connect you directly with the development team and provide a tailored briefing.

Order a report on this topic More in Digital & ICT