Europe’s high-performance computing (HPC) and artificial intelligence (AI) centres are powerful yet power-hungry. As demand for computing grows, so do the energy bill and ecological impact. SEANERGYS tackles this challenge by creating an integrated software (SW) suite to make operation of supercomputers more energy-efficient without sacrificing performance, and to help end-users to employ these resources in an energy-efficient way. The SW suite monitors energy use, predicts workload behaviour, and optimises job scheduling, allowing HPC/AI systems to do more with less. SEANERGYS aims to strike a balance between delivering compute results within a set energy budget and reducing energy use for the same output. The system is designed for production-level deployment up to Exascale, ensuring European HPC and AI centres run smarter, greener, and faster.
SEANERGYS creates an integrated European SW suite that optimises the operation of supercomputers. In doing so, it addresses four objectives: reducing the amount of energy used for real-world workload mixes as the primary objective, optimising resource utilisation, enhancing system throughput and reducing response time as secondary objectives. Since these can conflict with each other, site-specific policies will define the weights attached to each, and the SEANERGYS SW suite will tailor system operation towards the combined optimum. Scenarios include improving the throughput of HPC systems, generating more R&D results for a given energy budget, and producing a fixed set of R&D results with less energy, while striving to keep response times constant. The SEANERGYS SW solution consists of a comprehensive monitoring infrastructure (CMI), an Artificial Intelligence data analytics system (AIDAS), and a dynamic scheduling and resource management system (DSRM).
The CMI gathers data from hardware and SW sensors, and correlates it with scheduler information to identify jobs that do not fully utilise allocated resources. Users receive automatic feedback on energy and resource use for each run, plus actionable information on how to optimise these. The AIDAS leverages AI models trained with a vast set of operational data of the participating HPC/AI centres. It fingerprints resource usage patterns, predicts future job behaviour, and identifies complementary job profiles for potential co-scheduling. Finally, the DSRM utilises these insights to develop scheduling policies that maximise resource utilisation and energy efficiency, and supports workloads with dynamic and adaptable resource profiles.
The SEANERGYS suite will be ready for use at Exascale level. It builds on proven European projects, established expertise, and widely used open-source software to develop a SW suite that achieves the functionality, performance and stability needed by European HPC/AI centres, defined by KPIs and acceptance criteria and processes established at the project start. Development follows an agile, DevOps-based approach to provide full traceability by linking and tracking requirements, interface, functional and performance specifications, code design and development steps, and validation/verification throughout the development life cycle. Quality measures will include code reviews, automated SW quality analysis, unit and integration tests and a verification suite. The project will employ progressive testing and validation from single-node environments to mid-scale systems. Finally, there will be acceptance tests on production supercomputers..
