← Back to the Cloud and compute tiles
National and research compute
75 records Generated from this page's data. Click a name to open its panel on the Cloud and compute page. Records resting on one source are tagged "one source".| Record | What it is | Place | Links | Confidence |
|---|---|---|---|---|
| Ada, University of NottinghamData centres and compute sites | Nottingham's GPU-accelerated HPC cluster on 80+ Lenovo servers with NVIDIA A100 and H200 GPUs, delivered with OCF.university GPU clusterKey facts 80+ Lenovo ThinkSystem servers; NVIDIA A100 and H200 (counts not stated); four times the compute of a predecessor that served five years; OCF deployment and premier support; OCF case study dated 2025-07-29 | Nottingham | lenovo.com/sg/en/case-studies-customer-success-s · ocf.co.uk/about-us/latest-news/ · ocf.co.uk/about-us/latest-news/university-of-not | verified |
| AI Research Resource (AIRR)Schemes, programmes and AI lines · also on Data centres | The UK government's pool of public AI supercomputers, whose GPU time is given free to researchers and small UK firms.public compute | gov.uk/government/publications/ai-research-resou · gov.uk/government/publications/uk-compute-roadma · delivery.ai.gov.uk/2/ · gov.uk/government/publications/ai-opportunities- | verified | |
| AIRR Expansion AI Cloud Compute (£250m cloud GPU procurement)Schemes, programmes and AI lines | A planned government contract to rent cloud GPUs from a commercial provider, adding to the public AI Research Resource. plannedpublic compute | find-tender.service.gov.uk/Notice/081657-2025 · delivery.ai.gov.uk/2/ · find-tender.service.gov.uk/Notice/081657-2025/PD · ukauthority.com/articles/dsit-launches-procureme | verified | |
| AIRR heterogeneous national AI supercomputer (£750m)Schemes, programmes and AI lines · also on Chips and hardware | A planned £750m government AI supercomputer mixing established chips with new British chips, due to start in early 2028. plannedpublic compute | (host site not chosen) | gov.uk/government/publications/expression-of-int · gov.uk/government/news/a-decisive-shift-to-power · gov.uk/government/publications/uk-ai-hardware-pl · gov.uk/government/publications/digital-and-techn | verified |
| AlcesCompanies | Integration partner that helped the University of Manchester add NVIDIA H200 servers to its CSF3 research cluster.HPC integration partnerKey facts CSF3 expansion, Manchester (late 2025): four 8-way NVIDIA H200 HGX nodes (32 GPUs), 230 TB storage, Dell host servers with 2 × 32-core CPUs and 1.5 TB RAM, up to 11 kW each | research-it.manchester.ac.uk/news/2026/06/09/new | one source | |
| ARCHERData centres and compute sites | National supercomputer hosted at EPCC, University of Edinburgh, launched November 2013; lifetime cost £78.2m over five years; funded mainly by EPSRC and NERC. closed (replaced by ARCHER2) | Edinburgh (—) | ukri.org/wp-content/uploads/2022/07/EPSRC-050722 · bristol.ac.uk/acrc/high-performance-computing/na | verified |
| ARCHER2Data centres and compute sites | UK Tier-1 national supercomputer, an HPE Cray EX system at EPCC, University of Edinburgh; operational 2021; £79m from EPSRC and NERC; access ends November 2026. active; service ending | Edinburgh (—) | discover.ukri.org/archer2/index.html · docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · epcc.ed.ac.uk/hpc-services/archer2 · epcc.ed.ac.uk/sites/default/files/newsletters/EP | verified |
| ARCHIE-WeStData centres and compute sites | Regional Tier-2 HPC Centre of Excellence based at the University of Strathclyde, one of five EPSRC funded in 2012–13. | Glasgow (—) | ukri.org/wp-content/uploads/2022/07/EPSRC-050722 | one source |
| Atos (UK research HPC supplier)Companies | Supplier that delivered the JADE2 deep-learning supercomputer for Oxford and the DiRAC Tursa GPU system in Edinburgh.HPC and DGX integratorKey facts JADE2: four-year contract worth £5m with the University of Oxford; 63 NVIDIA DGX nodes with 504 V100; DGX SuperPOD design; hosted at STFC Hartree Centre (page undated); DiRAC Tursa, Edinburgh: ATOS named vendor; NVIDIA A100 40GB; 112 GPU nodes; committed 5.83 TF in tender, 8.8 TF sustained after commissioning; also JADE supplier in 2017 (DEAL-CLD-2019); the brief lists it as Atos/Eviden; UK legal entity not seen | oerc.ox.ac.uk/case-studies/atos-partners-with-un · dirac.ac.uk/extreme-scaling-service-edinburgh/ | verified | |
| AWE supercomputing, AldermastonData centres and compute sites | Compute facility of the Atomic Weapons Establishment at Aldermaston, listed in the UK Compute Roadmap annex. | Aldermaston (—) | gov.uk/government/publications/uk-compute-roadma | one source |
| BaskervilleData centres and compute sites | Tier-2 GPU facility at the University of Birmingham run with the Rosalind Franklin Institute, the Alan Turing Institute and Diamond Light Source. | Birmingham (—) | hpc-uk.ac.uk/facilities/ · ri.itservices.manchester.ac.uk/?p=7513 | verified |
| Baskerville GPU National Compute ResourceData centres and compute sites | New UKRI-funded GPU National Compute Resource at Birmingham, due to be available in 2027. planned | Birmingham (—) | ri.itservices.manchester.ac.uk/?p=7513 · datacenterdynamics.com/en/news/scotlands-datavit | one source |
| Bede (formerly NICE-19)Data centres and compute sites | EPSRC-funded Tier-2 GPU cluster run by N8 CIR at Durham: IBM Power9 nodes with Nvidia V100s plus GH200 Grace Hopper nodes; due for permanent decommissioning 31 July 2026. decommissioned (scheduled) | Durham (—) | docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · ri.itservices.manchester.ac.uk/?p=7513 · ocf.co.uk/higher-education | verified |
| BlueCrystal 4Data centres and compute sites | University of Bristol's own HPC cluster (Tier-3), used for test runs before applying to national services. | Bristol (—) | bristol.ac.uk/acrc/high-performance-computing/na | one source |
| BoxxeCompanies | Dell reseller awarded a £7.87m contract to deliver a modular HPC data centre with CPU and GPU nodes for UK Industrial Fusion Solutions.Dell reseller, modular HPCKey facts £7.87m contract from UK Industrial Fusion Solutions Ltd (UKAEA Group subsidiary; STEP programme), via the Southern University Purchasing Consortium HPC framework; Dell hardware; site likely Culham Science Centre; includes long-term support (DCD article, undated) | Companies House · datacenterdynamics.com/en/news/uk-fusion-energy- · find-and-update.company-information.service.gov. | one source | |
| Calder, University of LeedsData centres and compute sites | Leeds's next central HPC system, with 56 NVIDIA H200 GPUs on Dell servers supplied by MTI. purchase phase; due Spring 2026university GPU clusterKey facts 56 NVIDIA H200 141GB PCIe (7 nodes × 8); 98 CPU nodes of 768 GB; 12,228 AMD EPYC 9555 cores; 50 TB NFS and 5.4 PB Lustre; 100 Gb/s InfiniBand; predecessor Aire (84 GPUs, 9,072 cores); cost not published (2025-12) | Leeds | arc.leeds.ac.uk/knowledge-centre/calder-unveiled | one source |
| Cambridge Service for Data Driven Discovery (CSD3)Data centres and compute sites | Tier-2 service at the University of Cambridge combining the Cumulus CPU cluster and the Wilkes3 NVIDIA A100 GPU cluster, used for simulation, data analysis and AI. | Cambridge (—) | hpc-uk.ac.uk/facilities/ · docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou | verified |
| Charger (UCL-led National Compute Resource)Data centres and compute sites | UKRI-funded national computing resource led by UCL, built by HPE and hosted by DataVita in Lanarkshire; it follows the MMM Hub. being installed; expected later in 2026national CPU compute resourceKey facts £19.5m over five years (UKRI quote on same page: £19.4m), equipment plus service to 2031; more than 37,000 Intel Xeon Platinum CPU cores; HPE (Slingshot, Cray storage), direct liquid cooling; a small number of NVIDIA RTX 6000 Pro Blackwell server GPUs "in the cloud"; hosted at DataVita's Lanarkshire campus, about 465 t CO2e a year saved against London; one of four UKRI National Compute Resources | Lanarkshire (hosting); London (UCL ARC) | ucl.ac.uk/news/2026/feb/ucl-host-ps195m-supercom · datacenterdynamics.com/en/news/scotlands-datavit | verified |
| CirrusData centres and compute sites | HPC system at EPCC, University of Edinburgh, now one of the UKRI National Compute Resources; HPE EX4000 with 640 AMD EPYC nodes. | Edinburgh (—) | hpc-uk.ac.uk/facilities/ · docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · epcc.ed.ac.uk/sites/default/files/newsletters/EP · ed.ac.uk/news/step-change-for-uks-digital-horsep | verified |
| Comet, Newcastle UniversityData centres and compute sites | Newcastle University HPC system built with OCF on Lenovo servers and NVIDIA GPUs.university HPCKey facts Lenovo ThinkSystem servers, NVIDIA GPUs, Lustre parallel file system, managed with OCF Steel Stack (OCF case study 2026-03-09); counts and cost not seen (case-study page did not open); relation to Rocket (BIZ-CLD-2083) not stated | Newcastle | ocf.co.uk/about-us/latest-news/ | one source |
| CoSTAR GPU cluster, University of SurreyData centres and compute sites | GPU cluster with 100 NVIDIA H200s serving the CoSTAR creative-technology programme, hosted by the University of Surrey.creative-industries GPU clusterKey facts 100 NVIDIA H200 141 GB; 13 nodes (10 standard with 2 × AMD EPYC 9535; 2 Intel Xeon 8480+ pilot; 1 AMD pilot); 1,568 production cores; 750 TB WekaFS NVMe; InfiniBand NDR 400 Gb/s; funded through UKRI Infrastructure Fund, delivered by AHRC; partners Royal Holloway, Pinewood Studios, disguise, BT, Surrey County Council, Abertay University, National Film and Television School; vendor and go-live date not stated | England: South East | docs.pages.surrey.ac.uk/research_computing/hpc/c | one source |
| CSF3 NVIDIA H200 expansion, University of ManchesterData centres and compute sites | Manchester added 32 NVIDIA H200 GPUs to its Computational Shared Facility (CSF3) research cluster in late 2025. deployedGPU cluster expansionKey facts Four 8-way NVIDIA H200 HGX nodes (32 × H200 141 GB); 230 TB storage; Dell hosts, up to 11 kW each; Alces integration; deployed late 2025; Research Lifecycle Programme investment in CSF3 reported 2026-04-28; cost and funder not published | Manchester | research-it.manchester.ac.uk/news/2026/06/09/new · research-it.manchester.ac.uk/tag/csf/ | verified |
| DiRACSchemes, programmes and AI lines | STFC's HPC facility for theory and modelling in particle physics, astronomy, cosmology and nuclear physics, with services at Cambridge, Durham, Edinburgh and Leicester. | multi-site (—) | hpc-uk.ac.uk/facilities/ · docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · dirac.ac.uk/wp-content/uploads/2026/01/Wilkinson | verified |
| DiRAC COSMA (COSMA 8)Data centres and compute sites | DiRAC memory-intensive service at Durham University for cosmological simulations; the £10m COSMA 8 (Cosmology Machine 8), AMD-based, launched October 2023. | Durham (—) | docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · datacenterdynamics.com/en/news/durham-university · dirac.ac.uk/memory-intensive-durham/ | verified |
| DiRAC Data Intensive service, LeicesterData centres and compute sites | DiRAC data-intensive service at Leicester offering high-memory options; a second data-intensive DiRAC service runs at Cambridge. | Leicester (—) | docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · dirac.ac.uk/data-intensive-leicester/ | one source |
| DiRAC TursaData centres and compute sites | DiRAC's GPU-based extreme-scaling system at Edinburgh, used for workloads including quantum chromodynamics. | Edinburgh (—) | docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · dirac.ac.uk/extreme-scaling-service-edinburgh/ | one source |
| e-Infrastructure SouthData centres and compute sites | First-round (2012–13) regional Tier-2 HPC centre run by universities in the Science and Engineering South group. historical | England: South East | ukri.org/wp-content/uploads/2022/07/EPSRC-050722 | one source |
| Edinburgh exascale supercomputer plan (cancelled 2024)Schemes, programmes and AI lines | A planned national exascale computer at Edinburgh, cut in August 2024 and replaced in 2025 by an up-to-£750m national supercomputer. cancelled; replaced by BIZ-CLD-020exascale planKey facts proposed budget about £800m; University of Edinburgh built a £31m facility to house it; cancelled 2024-08-05; replacement up to £750m announced 2025-06; EPCC: "no longer calling this the exascale system" (2025-06-12) | Edinburgh | theregister.com/2024/08/05/axe_exascale_uk_edinb · theregister.com/2025/06/12/uk_national_supercomp | one source |
| EPCC additional storage for AI for scienceSchemes, programmes and AI lines | Government investment in more storage capacity at EPCC, Edinburgh, linked to the AIRR federated platform. announced | Edinburgh (—) | gov.uk/government/publications/ai-for-science-st | one source |
| EPCC Cerebras CS-3 clusterData centres and compute sites | Four Cerebras CS-3 wafer-scale systems at EPCC, University of Edinburgh, doubling its earlier two CS-2 systems; for AI model training and inference research. installed | Edinburgh (—) | datacenterdynamics.com/en/news/university-of-edi · epcc.ed.ac.uk/sites/default/files/newsletters/EP | one source |
| GridPPData centres and compute sites | Compute facility listed at Harwell in the UK Compute Roadmap's annex of UK compute facilities. | Harwell (—) | gov.uk/government/publications/uk-compute-roadma | one source |
| Hartree CentreData centres and compute sites · also on Networks and media | Compute centre at Daresbury listed in the UK Compute Roadmap annex; hosted the first JADE GPU service.Models and products (1)
| Daresbury (—) | gov.uk/government/publications/uk-compute-roadma · ukri.org/who-we-are/stfc/facilities/daresbury-la · hartree.stfc.ac.uk/news/2023/02/13/new-30-millio · hartree.stfc.ac.uk/wp-content/uploads/2024/04/Ha | one source |
| Hartree Centre, IBM and UKAEA fusion foundation-model partnershipPartnerships and contracts | STFC Hartree Centre, IBM and UKAEA will build AI foundation models from fusion experiment data to help design future powerplants. announcedfusion AI modelsKey facts Models to learn the dynamics of experimental fusion data; UKAEA contributes data and modelling from JET and MAST-U; no funding figure; named: Rob Akers (UKAEA), Kate Royse (Hartree), Juan Bernabe Moreno (IBM Research Europe, UK and Ireland), Vassil Alexandrov (Hartree) (2024-12-10) | Daresbury | hartree.stfc.ac.uk/news/2024/12/10/a-new-partner | one source |
| Hartree National Centre for Digital Innovation (IBM and STFC)Partnerships and contracts · also on Big tech in the UK | A five-year, £210m programme between IBM and the government's Hartree Centre at Daresbury to help UK businesses use AI and quantum computing. | Daresbury, Cheshire | uk.newsroom.ibm.com/2021-06-04-UK-STFC-Hartree-C · research.ibm.com/blog/ibm-stfc-accelerating-inno · ukri.org/news/stfc-launches-powerful-new-ai-supe · datacenterdynamics.com/en/news/447-petaflops-hpc | verified |
| HECToRData centres and compute sites | UK Tier-1 national HPC service from 2007, funded by EPSRC, NERC and BBSRC; lifetime cost £118m (2007 to 2013/14). closed | ukri.org/wp-content/uploads/2022/07/EPSRC-050722 | one source | |
| Higgs HPC cluster, Northumbria UniversityData centres and compute sites | Northumbria University's first university-wide HPC cluster, built on Lenovo servers with AMD EPYC processors and integrated by Logicalis.first university-wide HPCKey facts "Multi-million pound" (amount not disclosed); Lenovo ThinkSystem servers and storage, AMD EPYC 9004, Ubiquity software; capabilities to be reviewed and expanded over five years; GPU and node counts not stated (2024-11-20) | England: North East | newsroom.northumbria.ac.uk/pressreleases/superco · lenovo.com/gb/en/case-studies-customer-success-s | verified |
| HPC Midlands PlusNetworks, bodies, events and media | Midlands HPC consortium led by Loughborough University, funded as one of six Tier-2 centres in 2016; runs Sulis at Warwick. | Loughborough (—) | ukri.org/wp-content/uploads/2022/07/EPSRC-050722 · ri.itservices.manchester.ac.uk/?p=7513 | verified |
| HX2, Imperial College LondonData centres and compute sites | £10m Imperial supercomputer from Lenovo and Intel with NVIDIA H200 GPU nodes, designed under the ICICLE partnership. being deployed (2025)university GPU clusterKey facts £10m; 24 Lenovo SD650-N v3 GPU nodes, each 2 × Intel Xeon 8562Y and 4 × NVIDIA H200 (96 H200 by node arithmetic); 96 Lenovo SC750 V4 CPU nodes, each 2 × Xeon 6960; 257 TB RAM; 1,935 TB usable storage; first UK academic deployment of SC750 V4 Neptune; unnamed London colocation site | London | datacenterdynamics.com/en/news/intel-and-lenvo-p | one source |
| ICICLE (Imperial College, Intel and Lenovo HPC partnership)Partnerships and contracts | Multi-year Imperial, Intel and Lenovo collaboration on HPC and AI; its first direct liquid-cooled cluster went live at Digital Realty, Woking.liquid-cooled university HPCKey facts Announced 2023-05-16: Lenovo Neptune liquid cooling, Intel High Bandwidth Memory, training and workshops; liquid-cooled cluster contract signed 2025-07, handed over Q4 2025 at Digital Realty LGW14, Woking (campus 200,000 sq ft; racks over 100 kW); hardware of that cluster not disclosed; HX2 designed under ICICLE | London; Woking (hosting) | news.lenovo.com/pressroom/press-releases/icicle- · datacenterdynamics.com/en/news/uks-imperial-coll | verified |
| IRISData centres and compute sites | Research computing infrastructure listed at Harwell in the UK Compute Roadmap's annex of UK compute facilities. | Harwell (—) | gov.uk/government/publications/uk-compute-roadma | one source |
| Isambard (first GW4 system)Data centres and compute sites | Arm-based Tier-2 supercomputer funded by EPSRC for the GW4 universities (Bristol, Bath, Cardiff, Exeter), led by the University of Bristol. historical (succeeded by Isambard 3) | Bristol (—) | bristol.ac.uk/acrc/high-performance-computing/na · ukri.org/wp-content/uploads/2022/07/EPSRC-050722 · bristol.ac.uk/research/centres/bristol-supercomp | verified |
| Isambard 3Data centres and compute sites | Tier-2 supercomputer of the GW4 universities led by the University of Bristol; 384 NVIDIA Grace CPU Superchip nodes; separate from Isambard-AI. | Bristol (—) | ukri.org/councils/epsrc/facilities-and-resources · docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · bristol.ac.uk/news/2023/may/isambard-3-investmen | verified |
| Isambard-AI data repository (BriCS)Schemes, programmes and AI lines | New government data repository to sit alongside Isambard-AI at the Bristol Centre for Supercomputing. announced | Bristol (—) | gov.uk/government/publications/ai-for-science-st | one source |
| JADE (Joint Academic Data science Endeavour)Data centres and compute sites | Oxford-led deep-learning Tier-2 service at the STFC Hartree Centre, Daresbury: 22 Nvidia DGX-1 servers supplied by Atos; £3m from EPSRC (2017); decommissioned September 2021. decommissioned | Daresbury (—) | hpc-uk.ac.uk/facilities/ · datacenterdynamics.com/en/news/uk-universities-g | verified |
| JADE 2.5Data centres and compute sites | Current JADE GPU service, listed by EPSRC among Tier-2 services researchers can apply to use. | ukri.org/councils/epsrc/facilities-and-resources | one source | |
| JADE IIData centres and compute sites | 2020 renewal of JADE: 63 Nvidia DGX MAX-Q servers with eight V100 GPUs each; compute withdrawn November 2024, all access ended 6 January 2025. retired | docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · services.ncl.ac.uk/itservice/research/rescomp · oerc.ox.ac.uk/case-studies/atos-partners-with-un | verified | |
| JASMINData centres and compute sites | NERC facility for environmental data analysis, listed at Harwell, Oxfordshire, in the UK Compute Roadmap annex. | Harwell (—) | gov.uk/government/publications/uk-compute-roadma · ukri.org/councils/nerc/facilities%20and%20resour | verified |
| JOINER network GPU platformsSchemes, programmes and AI lines | Three JOINER network terminals, in Leeds, Bristol and Edinburgh, awarded new Nvidia GPU platforms. awarded | Leeds; Bristol; Edinburgh (—) | gov.uk/government/publications/digital-and-techn | one source |
| Logicalis UKICompanies | IT services firm that integrated Northumbria University's Lenovo-based Higgs HPC cluster and works with the university for five years.HPC integratorKey facts Integrator for Higgs, Northumbria's first university-wide HPC: Lenovo ThinkSystem, AMD EPYC 9004, Ubiquity software; "multi-million pound" (amount not disclosed); five-year engagement (2024-11-20); CEO Logicalis UKI Neil Eke | Companies House · newsroom.northumbria.ac.uk/pressreleases/superco · lenovo.com/gb/en/case-studies-customer-success-s · find-and-update.company-information.service.gov. | verified | |
| Lovelace (Loughborough University HPC)Data centres and compute sites | Loughborough University's £1m supercomputer built by OCF on Lenovo servers, with a small set of NVIDIA V100 GPU nodes.university GPU clusterKey facts £1m (2021-08); 58 Lenovo SD530 compute nodes, three GPU nodes each with two NVIDIA Tesla V100, two visualisation nodes; 800TB Lenovo DSS-G storage (2021-08) | Loughborough | ocf.co.uk/about-us/blog/ocf-works-with-loughboro | one source |
| Mary Coombs (Hartree Centre AI supercomputer)Data centres and compute sites | 24.41-petaflop GPU-based AI supercomputer run by STFC's Hartree Centre at Sci-Tech Daresbury, in a new £30m supercomputing centre. deployed | Sci-Tech Daresbury (—) | datacenterdynamics.com/en/news/uk-stfc-deploys-2 · hartree.stfc.ac.uk/news/2025/10/20/stfc-launches | one source |
| Materials and Molecular Modelling Hub (Young cluster)Data centres and compute sites | Tier-2 computing for materials and molecular modelling, led by UCL for eight partners; its Young cluster lists 23,040 CPU cores and 48 NVIDIA A100 GPUs. | London (—) | hpc-uk.ac.uk/facilities/ · docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · ucl.ac.uk/news/2026/feb/ucl-host-ps195m-supercom | verified |
| Maxwell (University of Aberdeen HPC)Data centres and compute sites | University of Aberdeen research cluster for genomics, designed, integrated and managed by OCF on Lenovo servers with NVIDIA GPUs.university GPU clusterKey facts 40 compute nodes plus four Lenovo SD530 servers; NVIDIA GPUs (count not published); BeeGFS storage, "twenty times more storage" than predecessor (2020-08) | Aberdeen | ocf.co.uk/about-us/blog/maxwell-supercomputer-at | one source |
| Met Office supercomputer (Microsoft Azure-based)Data centres and compute sites · also on Big tech in the UK | Met Office forecasting supercomputer run by Microsoft on Azure in dedicated halls of Microsoft data centres under a £1.2bn contract awarded February 2021. | Exeter (operator HQ; system location not stated) (—) | gov.uk/government/news/record-breaking-150bn-inv · datacenterdynamics.com/en/news/met-office-switch | verified |
| Monsoon2 (joint NERC–Met Office service)Data centres and compute sites | Joint NERC and Met Office supercomputing service on the Met Office Cray XC40, for environmental research. | Exeter (—) | ukri.org/wp-content/uploads/2025/09/NERC-150925- · ukri.org/councils/nerc/facilities%20and%20resour | verified |
| MTICompanies | Company that won the University of Leeds tender for its Calder HPC system, supplying Dell hardware.HPC reseller (Dell)Key facts Calder, Leeds (2025-12): 7 GPU nodes × 8 NVIDIA H200 141GB = 56 GPUs; 98 CPU nodes; 12,228 AMD EPYC 9555 cores; 5.4 PB Lustre; service and support from Cambridge Research Computing; target operational Spring 2026 | arc.leeds.ac.uk/knowledge-centre/calder-unveiled | one source | |
| N8 CIR (N8 Centre of Excellence in Computationally Intensive Research)Networks, bodies, events and media | Consortium of eight northern English universities (Durham, Lancaster, Leeds, Liverpool, Manchester, Newcastle, Sheffield, York), formerly N8 HPC, which runs Bede. | multi-site (—) | ukri.org/wp-content/uploads/2022/07/EPSRC-050722 · hpc-uk.ac.uk/facilities/ | verified |
| Next National Supercomputing Service, University of EdinburghSchemes, programmes and AI lines | A planned £750m national supercomputer at the University of Edinburgh, to replace ARCHER2 and serve science and AI research. plannedpublic compute | Edinburgh | gov.uk/government/news/scotland-to-host-uks-nati · gov.uk/government/publications/uk-compute-roadma · find-tender.service.gov.uk/Notice/062921-2025 · gov.uk/government/publications/digital-and-techn | verified |
| NI-HPC (Kelvin2)Data centres and compute sites | Northern Ireland's Tier-2 HPC centre run by Queen's University Belfast and Ulster University; Kelvin2 includes NVIDIA V100/A100/H100, AMD MI300X and Intel GPUs. | Belfast (—) | hpc-uk.ac.uk/facilities/ · docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou | verified |
| PitchwiseCompanies | University of Bristol student spin-out making AI video tracking for grassroots football; awarded 5,500 GPU hours on Isambard-AI.AIRR compute recipientKey facts 5,500 GPU hours on Isambard-AI valued at £25,000, through UKRI's AI Research Resource, supported by Innovate UK (2025-11-06); phone-mounted hardware, AI vision software and companion app for match tracking, highlights and stats | Bristol | bristol.ac.uk/news/2025/november/spin-out-superc | one source |
| Quantinuum–Hartree Centre quantum computing accessPartnerships and contracts | Quantinuum makes its H1-Series quantum computers available to UK businesses and researchers through the STFC Hartree Centre. announcedquantum hardware accessKey facts H1-Series hardware via cloud or on site at Hartree; target areas quantum chemistry, computational biology, quantum AI, quantum-augmented cybersecurity; terms not disclosed; Quantinuum CEO Raj Hazra (2024-07-11) | Daresbury | hartree.stfc.ac.uk/news/2024/07/11/partnership-b | one source |
| Quantum Dice–Hartree Centre collaborationPartnerships and contracts | Oxford quantum spin-out Quantum Dice and the Hartree Centre agreed to widen industrial use of its random-number and probabilistic computing technology. signedQRNG, probabilistic computingKey facts Collaboration agreement plus Joint Statement of Endeavour (2025-12-03); short term: QRNG use cases in materials, manufacturing, product design, logistics; longer term: validate the probabilistic computing platform and host it at Hartree; no financial terms; CEO Ramy Aboushelbaya, COO Zhanet Zaharieva | Daresbury | hartree.stfc.ac.uk/news/2025/12/03/stfc-hartree- | one source |
| Rocket (Newcastle University HPC)Data centres and compute sites | Newcastle University's university-wide HPC service with over 5,000 cores and 500TB storage (Tier-3). | Newcastle upon Tyne (—) | services.ncl.ac.uk/itservice/research/rescomp | one source |
| Scafell Pike (Hartree Centre)Data centres and compute sites | Hartree Centre system using Intel Skylake CPUs and Xeon Phi accelerators, due to be replaced by a 44.7-petaflop Lenovo system. in service March 2024; due for replacement | Daresbury (—) | datacenterdynamics.com/en/news/447-petaflops-hpc · news.lenovo.com/pressroom/press-releases/stfc-ha | one source |
| SCARFData centres and compute sites | Compute service that university HPC teams suggest for researchers within the STFC remit. | ri.itservices.manchester.ac.uk/?p=7513 | one source | |
| Sovereign AI AIRR launch opportunity (call for researchers)Schemes, programmes and AI lines | An early government call giving UK research groups short projects on the Isambard-AI supercomputer. supersededpublic compute | gov.uk/government/publications/sovereign-ai-airr · gov.uk/government/publications/digital-and-techn · gov.uk/government/publications/ai-for-science-st | one source | |
| StackHPCCompanies | UK small firm supplying the software stack for Cambridge's Dawn AI supercomputer, with Dell; CEO John Taylor. | gov.uk/government/news/technology-secretary-anno · gov.uk/government/news/cambridge-supercomputer-s · stackhpc.com/ · stackhpc.com/cambridge-dawn.html | verified | |
| Stanage, University of SheffieldData centres and compute sites | Sheffield's central HPC cluster on Dell servers, with 104 NVIDIA GPUs stated, including newly added H100 NVL nodes. active; GPU nodes added 2025-12 and 2026-01university GPU clusterKey facts Stated totals: 104 GPUs, 205 worker nodes, 12,896 CPU cores (per-node rows on the page do not sum to these); 13 Dell XE8545 nodes with 4 × A100 80GB; 6 Dell R7525 nodes with 2 × H100 80GB; Dell R760xa nodes with 4 × H100 NVL 94GB, 16 H100 NVL after two more nodes (2026-01); 2.2 PB Lustre; one 8 × RTX 6000 Pro server ordered (2026-01); cost not published | Sheffield | docs.hpc.shef.ac.uk/en/latest/stanage/cluster_sp · docs.hpc.shef.ac.uk/en/latest/newsletter/2026011 | verified |
| SulisData centres and compute sites | Tier-2 platform for ensemble computing hosted at the University of Warwick for the HPC Midlands+ consortium; UKRI funding and access routes ended March 2026. UKRI funding ended | Coventry (—) | hpc-uk.ac.uk/facilities/ · docs.hpc.shef.ac.uk/en/latest/other-uk-hpc-resou · ocf.co.uk/portfolio/sulis-brings-dedicated-high- | verified |
| Topsy (Newcastle SAgE Faculty HPC)Data centres and compute sites | 1,000-core HPC system owned and run by Newcastle University's Science, Agriculture and Engineering (SAgE) Faculty. | Newcastle upon Tyne (—) | services.ncl.ac.uk/itservice/research/rescomp | one source |
| UKAEA fusion energy supercomputerData centres and compute sites | £45m supercomputer for fusion energy research by the UK Atomic Energy Authority, due to launch later in 2026. planned | gov.uk/government/publications/digital-and-techn | one source | |
| Unilever–Hartree Centre computing partnershipPartnerships and contracts | Unilever R&D has used the STFC Hartree Centre as a materials chemistry computing partner for over a decade, now including AI and quantum.industrial HPC partnershipKey facts Partnership "for over a decade" (Hartree Centre Strategy 2024–2029, published 2024-04); covers several science areas, talent, and joint AI and quantum work; quote attributed to Alberto Prado (role not captured); money not stated | Daresbury | hartree.stfc.ac.uk/wp-content/uploads/2024/04/Ha | one source |
| University of Southampton Lenovo H200 HPC systemData centres and compute sites | Southampton's new Lenovo-built HPC with NVIDIA H200 GPUs, the first order (about £7m) under a four-year supplier framework. ordered; delivery summer 2026university GPU clusterKey facts Initial order ≈ £7m, delivery summer 2026; Lenovo ThinkSystem SR675 V3 with NVIDIA H200 and NVLink; phase 2 expected on NVIDIA Grace Blackwell with Lenovo ThinkSystem SC777 V4 Neptune (no date); four-year framework after competitive tender; aim to re-enter the TOP500; GPU count not published | Southampton | southampton.ac.uk/news/2026/06/new-partnership-t · news.lenovo.com/pressroom/press-releases/lenovo- | verified |
| Viking (University of York)Data centres and compute sites | University of York's research compute cluster (Tier-3), accessed through an OpenOnDemand portal. | York (—) | york.ac.uk/it-services/tools/research/research-i | one source |
| Zenith, University of CambridgeData centres and compute sites | A public AI supercomputer at the University of Cambridge, built with AMD, Dell and StackHPC, that forms part of the AI Research Resource.public compute | Cambridge | gov.uk/government/publications/ai-research-resou · gov.uk/government/publications/digital-and-techn · gov.uk/government/news/cambridge-supercomputer-s | one source |