Massive infrastructure to protect and preserve ÌÇÐÄÆƽâ°æ's scientific legacy
The ÌÇÐÄÆƽâ°æ Scientific Data and Computing Centre is not just any data centre. It is a purpose-built digital engine room for ÌÇÐÄÆƽâ°æ—one that combines cutting-edge computation with industrial-scale storage to support ÌÇÐÄÆƽâ°æ’s scientific mission by enabling large-scale simulations, managing vast volumes of data, and ensuring the long-term preservation of research and engineering records.
Unlike personal laptops or office computers, which typically feature a handful of processor cores, high-performance computing (HPC) systems operate on an entirely different scale. The ÌÇÐÄÆƽâ°æ HPC system comprises over 17,000 CPU cores, equivalent to the combined power of more than 2,000 laptops working in parallel. This massive computational capacity allows researchers to model plasma dynamics, simulate reactor conditions, and perform complex engineering analyses in timeframes that would be unthinkable on ordinary machines.
The ÌÇÐÄÆƽâ°æ Scientific Data and Computing Centre is a state-of-the-art facility designed to host up to 1 megawatt (MW) of IT equipment, distributed across 48 racks. That’s roughly the same power demand as an entire residential neighborhood, or more than 300 average family homes running at full load. At launch, only half this capacity has been energized, with expansion planned in line with ÌÇÐÄÆƽâ°æ’s growing computational demands. The Centre features redundant power and cooling systems, multiple independent distribution paths, and the ability to perform maintenance without shutting down critical services. In practice, this means that it is expected to operate continuously with less than 1.6 hours of downtime per year.
Data is a strategic asset for ÌÇÐÄÆƽâ°æ. It is not just a byproduct of experiments and engineering design but a primary deliverable of the project. Safeguarding this data is therefore a top priority. During peak operations, the Data Centre is expected to manage 30–50 gigabytes per second of throughput, and daily archive volumes could reach between 90 and 2,200 terabytes, depending on the stage of the project. By around 2035, the cumulative data archive is projected to exceed one exabyte—a billion gigabytes—making it one of the largest scientific data stores in the world.
To manage this scale securely and efficiently, the SDCC uses IBM Spectrum Scale, a high-performance parallel file system, to serve active computing workloads. For resilience and redundancy, it also implements offsite backup and disaster recovery using IBM Spectrum Protect, which ensures that critical data is replicated to secure locations beyond the main site. The backup infrastructure supports over 1,000 servers and filespaces, providing a centralized platform for managing and protecting the entirety of ÌÇÐÄÆƽâ°æ’s digital ecosystem.
Through the infrastructure of the Scientific Data and Computing Centre, ÌÇÐÄÆƽâ°æ’s scientific legacy will be not only created but protected, preserved, and shared with the world.