EDRA · Project

Cloud Service That Speeds Up Heavy Computing on Amazon Web Services

digitalPilotedTRL 7

Imagine you have a really complex calculation to run — like simulating weather or testing a car design — and your regular computer just can't keep up. A previous EU project built special hardware that reshapes itself on the fly to match whatever task you throw at it, running things much faster. EDRA took that technology and packaged it as a ready-to-use service on Amazon's cloud, so any company can rent that computing power without buying expensive hardware. You just pick an "EDRA-enhanced" virtual machine on AWS, upload your application, and it automatically runs at peak performance for less money.

By the numbers

EUR 100,000

EU funding for commercialization activities

Cloud platform targeted (AWS marketplace)

Total project deliverables produced

18 months

Project duration (May 2019 to October 2020)

The business problem

What needed solving

Companies running heavy computational workloads — simulations, data analytics, machine learning training — face a painful trade-off: either invest millions in on-premise HPC hardware or pay steep cloud computing bills for generic virtual machines that aren't optimized for their tasks. Standard cloud instances waste processing power because the hardware can't adapt to different workload patterns, leaving performance and money on the table.

The solution

What was built

The project produced 3 deliverables including an AWS marketplace release of EDRA-enhanced virtual machines and an IP exploitation report mapping out commercial partnerships. The core product is a cloud service that automatically maps user applications onto reconfigurable hardware within Amazon's infrastructure, built on technology from the earlier EXTRA exascale computing project.

Audience

Who needs this

Cloud-native companies with compute-intensive workloads on AWSEngineering firms running simulations (CFD, FEA, structural analysis)Financial institutions doing quantitative modeling and risk analysisResearch organizations needing affordable HPC without on-premise clustersAI/ML companies looking to accelerate training workloads on cloud infrastructure

Business applications

Who can put this to work

Cloud Computing & HPC Services

any

Target: Companies running large-scale simulations or data processing on AWS

If you are a technology company spending heavily on cloud computing for intensive workloads like simulations or analytics — this project developed a specialized virtual machine image for AWS that automatically maps your applications onto reconfigurable hardware for optimal performance and minimal cost. The tool was built specifically for the Amazon Web Services marketplace, making adoption straightforward.

Financial Services

mid-size

Target: Fintech firms running risk modeling and quantitative analysis

If you are a financial firm running computationally expensive risk models, pricing algorithms, or fraud detection — this project created a cloud-based acceleration service on AWS that can speed up number-crunching workloads. Because it deploys as a standard virtual machine on Amazon's infrastructure, your existing cloud operations team can test it without new hardware purchases.

Engineering & Manufacturing Simulation

enterprise

Target: Engineering firms using computational fluid dynamics or structural analysis

If you are an engineering company running complex simulations for product design or testing — this project packaged high-performance reconfigurable computing as an AWS marketplace service. Instead of investing in on-premise HPC clusters, your engineers can spin up EDRA-enhanced virtual machines on demand and only pay for the compute time they actually use.

Frequently asked

Quick answers

What would this cost my company to use?

EDRA was designed as an AWS marketplace service, so pricing would follow a pay-per-use cloud model. The project itself received EUR 100,000 in EU funding as a Coordination and Support Action focused on commercialization. Specific end-user pricing details are not available in the project data.

Can this handle enterprise-scale workloads?

The underlying technology comes from the EXTRA project, which targeted exascale High Performance Computing applications — the highest tier of computing power. By deploying on Amazon Web Services infrastructure, the solution is designed to scale with AWS's existing cloud capacity. However, the project had only 1 partner and a limited budget, so enterprise validation at scale would need to be confirmed.

Who owns the IP and how is it licensed?

The IP is held by the Greek research institute EREVNITIKO PANEPISTIMIAKO INSTITOUTO TILEPIKONONIAKON SYSTIMATON. One of the 3 deliverables was specifically an 'EDRA IP exploitation report' covering potential collaborations with companies to enhance market penetration. Licensing terms would need to be discussed directly with the coordinator.

Is this actually available on AWS right now?

The project deliverable D3.3 described the first EDRA release on the AWS marketplace. The project closed in October 2020. Based on available project data, current availability on the AWS marketplace should be verified directly, as cloud offerings can change after project funding ends.

What makes this different from other cloud HPC solutions?

EDRA uses a Decoupled Access-Execute approach with reconfigurable hardware, meaning the hardware itself adapts to your specific workload rather than using fixed-purpose processors. This comes from the EXTRA project's work on exascale computing. The key differentiator is automatic application mapping — users don't need to manually optimize their code for the hardware.

Does my team need special skills to use this?

The project specifically aimed at automatic mapping of applications to the reconfigurable hardware. End-users should be able to deploy applications onto EDRA-enhanced virtual machines without deep hardware expertise. However, the solution was designed within the AWS ecosystem, so familiarity with Amazon's cloud infrastructure is expected.

Consortium

Who built it

This is a single-partner project run entirely by one Greek university research institute, with no industry partners, no SMEs, and zero private-sector involvement. While the academic team clearly has deep technical expertise from the predecessor EXTRA project, the absence of any commercial partner is a notable gap for a project focused on market deployment. The EUR 100,000 budget is small, consistent with a Coordination and Support Action rather than full product development. A business looking to adopt this technology would be dealing directly with an academic institution, which may affect responsiveness, support expectations, and long-term product maintenance.

EREVNITIKO PANEPISTIMIAKO INSTITOUTO TILEPIKONONIAKON SYSTIMATONCoordinator · EL

How to reach the team

Greek university research institute specializing in telecommunications systems — contact through project website or CORDIS portal

Order a report on this consortium See EREVNITIKO PANEPISTIMIAKO INSTITOUTO TILEPIKONONIAKON SYSTIMATON profile →

Next steps

Talk to the team behind this work.

Want to explore whether EDRA's cloud-based HPC acceleration fits your computing workloads? SciTransfer can arrange a direct introduction to the research team and help evaluate the business case for your specific use.

Order a report on this topic More in Digital & ICT