We have un-matched partnership networks with world-class tech vendors.
Deep Green
Deep Green is a UK company reimagining where and how compute happens: it deploys small, high-density data centres inside buildings that need heat whether they be swimming pools, district heat networks, industrial sites. It gifts the waste heat from the dat centre, slashing both carbon footprint and energy cost.
HPC's dirty secret is heat: nearly all electrical input becomes thermal output, usually rejected expensively to atmosphere. Deep Green's immersion- and liquid-cooled edge data centres capture up to ~96% of that heat and reuse it, turning a cost centre into community and commercial value.
For research organisations under sustainability mandates (and UK universities have aggressive net-zero commitments), hosting batch HPC or AI workloads on Deep Green infrastructure delivers verifiable carbon and cost benefits. Because the heat is genuinely reused, the effective PUE and carbon accounting are among the best available anywhere.
Deep Green's model suits compute-heavy workloads; paired with OCF-designed storage and data-movement strategies, it can host processing close to renewable/reuse contexts while primary data remains at the institution.
AI training is precisely the kind of dense, interruptible/batch, GPU-heavy workload that fits distributed heat-reuse sites: high utilisation, high thermal output, and flexibility about location. Deep Green enables organisations to grow AI capacity without expanding on-campus data centre footprint or power envelope.
Deep Green turns HPC/AI waste heat into a genuine sustainability and cost advantage aligning it with the UK's decarbonisation agenda.
