UKRI Living Benchmarks
UKRI Living Benchmarks is a community‑driven initiative to turn skills for Digital Research Infrastructure (DRI) into measurable, reusable, and continually updated (“living”) benchmarks. Its goal is to help researchers and research technical professionals make the most of UK investments in HPC, cloud, data platforms, AI/ML, and research software by tying skills to real, end‑to‑end research tasks and observable outcomes—such as performance, scalability, cost‑efficiency, reproducibility, and FAIR/data stewardship. By focusing on practical, role‑relevant capabilities, the project creates a common language for what “good” looks like across disciplines and facilities.
The project delivers an openly available catalogue of benchmark scenarios, datasets, workflows, and assessment rubrics that mirror authentic research activities—from data ingress and curation through workflow orchestration, optimisation, and verification. Materials are designed to be portable and automatable where possible, enabling use across sites and systems, and to align with existing good practice (e.g., versioning, containers, provenance, security). Guidance is provided for training providers and facilities to adopt and adapt the benchmarks, integrate them into courses and onboarding, and gather consistent evidence on learner progress and service readiness. A contribution model invites the community to propose, review, and maintain benchmarks so they evolve with technology and user needs.
Outcomes include a shared evidence base for targeting skills investment, clearer progression pathways and recognition for research technical professionals, and faster, more effective uptake of DRI services. Regular releases keep the benchmark suites current, while an emphasis on openness, accessibility, and EDI supports reuse across domains and institutions. The net effect is to translate infrastructure spending into tangible gains in productivity, reliability, and research quality.