Scalable HPC infrastructure on AWS
Cost efficiency and stability through modernization
Qoro Quantum UG, headquartered in London, UK, develops innovative simulation software for quantum processors. The company enables complex algorithms to be efficiently emulated and tested on classical high-performance computing (HPC) clusters before they are deployed on real quantum hardware.
The project at a glance
- Zero-touch operations: Fully automated scaling eliminates manual maintenance almost entirely.
- Approximately 40% cost reduction through the intelligent use of AWS Spot Instances instead of static servers.
- Fail-safe during peak loads: The architecture automatically handles unexpectedly high request volumes without compromising performance.
- No investment risk for the customer thanks to full financing via AWS PoC Funding.
Background
Qoro Quantum faced a typical challenge: its existing infrastructure was largely deployed and scaled manually, leading to the potential of the public cloud not being used efficiently. This resulted in a lack of standardization and unnecessary fixed costs, as servers had to be paid for even when no simulations were running. Another technical problem was the synchronous communication of services during long-running simulations. Scalability was moreover limited by manual intervention. In order to reliably serve enterprise customers, Qoro Quantum needed a robust solution that ran stably and only incurred costs when actually used.
Solution
After meeting at the AWS Summit in London, Qoro Quantum approached codecentric to optimize their platform technologically. The implementation took place within a compact timeframe of around two months. During this time, two cloud specialists from codecentric worked closely with Qoro Quantum's lead software architect. Together, the project team developed an architecture that meets both the high performance requirements of the simulation workloads and the goals of scalability and cost efficiency.
Instead of merely implementing a technical optimization, codecentric used the project to strategically and technologically future-proof Qoro Quantum's platform. A decisive advantage for budget planning was the use of AWS Proof-of-Concept funding, which meant that Qoro Quantum did not incur any direct project costs.
Various alternative solutions were evaluated in advance. Classic deployments on virtual machines were rejected due to the high operating costs. Completely serverless approaches were ruled out because they could not adequately meet the specific requirements of the computationally intensive workloads. The chosen architecture thus proved to be the optimal compromise between flexibility, degree of automation, and cost control.
The solution was based on two strategic pillars:
Cost efficiency through pay-per-use: The platform was designed so that computing resources are only used when simulations are actually being run. This is made possible by the use of short-lived, cost-effective cloud resources that can be flexibly scaled up and down. As a result, costs are incurred almost exclusively when the resources are actually used.
Automation instead of manual administration: The infrastructure automatically responds to peak loads and provides additional resources as needed – these are shut down again when there is no activity. Through the consistent use of infrastructure as code, deployments could be made transparent, reproducible, and secure. This significantly reduced the operational effort required for operation and maintenance.
Result
The project has significantly strengthened Qoro Quantum's operational base. The new platform not only runs more stably, but is also noticeably cheaper to operate. The cost savings of around 40% enable the company to invest resources more specifically in software development.
The new elasticity is particularly valuable: even when an unexpectedly high number of requests arrive simultaneously, the system remains stable and reliably processes the jobs – a scenario that would have led to delays in the past. Technically, a low-maintenance environment has been created that takes the pressure off the team.
Thanks to the partnership with codecentric and the use of AWS funding programs, Qoro Quantum was able to carry out this transformation without financial risk and now has a platform that is ready for further growth.
Stephen DiAdamo
CTO, Qoro Quantum
Thanks to codecentric, we were able to greatly reduce our cloud costs and move to an infrastructure that scales automatically. What used to require manual effort is now much more stable and predictable, and AWS funding allowed us to make this change without upfront cost.
Stephen DiAdamo
CTO, Qoro Quantum
Any questions about the project?
Does your IT department also incur unnecessary costs, or do you have problems with scaling? Feel free to contact me!
IT Consultant
Further reference projects
Find out about other successful projects that we have completed with our customers. Perhaps you will find inspiration for a use case in your company here.
Florian Lüdiger
IT Consultant