News
DealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory鈥檚 NVIDIA & AMD Enterprise ExpansionDealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory鈥檚 NVIDIA & AMD Enterprise Expansion
EveryNeocloud
The north face of the National Computational Infrastructure building at the ANU

National Computational Infrastructure

Australia's national high-performance computing and data facility at ANU in Canberra

Photo: Nick-D, CC BY-SA 3.0, via Wikimedia Commons

The National Computational Infrastructure is a high-performance computing and data services facility at the Australian National University in Canberra, backed by the Australian Government's NCRIS programme and a collaboration of science agencies and universities. Its Gadi cluster and Tenjin partner cloud serve researchers across Australia, with access granted through partner funding and the National Computing Merit A

Share on WhatsApp

  • Gadi cluster rated at 9.26 PetaFLOP with 145,152 Intel cores
  • 160 Gadi nodes each carry four Nvidia V100 GPUs
  • HDR Mellanox InfiniBand Dragonfly+ networking up to 200 Gbit/s
  • 117 petabytes across fast storage, large files and tape archive
  • Descends from ANU Supercomputing Facility, running since 1987

What it is

NCI Australia runs high-performance computing and data services from the Australian National University campus in Canberra. Funding comes from the Australian Government's National Collaborative Research Infrastructure Strategy, with operational support from CSIRO, the Bureau of Meteorology, Geoscience Australia, the Australian Research Council, ANU and research-intensive universities and medical institutes. Machine time goes to these partners and to researchers who win grants through the National Computing Merit Allocation Scheme. Its main systems are the Gadi cluster and the smaller Tenjin partner cloud. NCI also holds large research datasets and provides storage and data services to Australian science.

How it came about

NCI grew out of the ANU Supercomputing Facility, which ran from 1987 to 1999. Around 2000 the Australian Government set up the Australian Partnership for Advanced Computing, and the APAC National Facility was placed at ANU and commissioned in 2001. APAC began changing into the present NCI collaboration in 2007. Earlier machines include Vayu, a Sun Microsystems cluster launched in November 2009 and fully commissioned in March 2010, which was later replaced by Raijin. Directors have included Lindsay Botten, Sean Smith and, currently, Andrew Rohl.

The compute

Gadi is a distributed-memory cluster rated at 9.26 PetaFLOP, built from 3,024 nodes holding 145,152 Intel Xeon Scalable 'Cascade Lake' cores, with 567 terabytes of main memory. A further 160 nodes carry four Nvidia V100 GPUs each. Networking uses HDR Mellanox InfiniBand in a Dragonfly+ topology, giving transfers up to 200 Gbit/s. Tenjin is a 67 TeraFLOP partner cloud with 1,600 Intel Xeon Sandy Bridge cores, 25 terabytes of memory and 160 terabytes of solid-state disk. The earlier Vayu system ran CentOS with Lustre storage and a variant of OpenPBS for job queues.

Where the GPUs are

The facility sits on the ANU campus in Canberra, in a building that uses hot aisle containment and free cooling to keep the computers at working temperature. Storage is tiered: 20 petabytes for fast input and output, 47 petabytes for large research files, and 50 petabytes on tape for archives. NCI describes these as the fastest filesystems in the Southern Hemisphere. Partner organisations include the Bureau of Meteorology, CSIRO and Geoscience Australia.

National Computational Infrast
A drawing made by EveryNeocloud from what this page says. It is a diagram, not a picture of the real provider.

Customers and deals

Access is shared between the funding partners and researchers who win time through the National Computing Merit Allocation Scheme. Named partners include CSIRO, the Bureau of Meteorology, Geoscience Australia, the Australian Research Council, ANU, plus research universities and medical institutes. Work carried out on the systems covers the Southern Sky Survey with the SkyMapper telescope, the Australian Community Climate and Earth System Simulator, COVID-19 mitigation studies, and medical and materials research.

Money and ownership

The Australian Government provided around A$26m to build the centre and install Vayu. ANU and CSIRO each subscribed about A$3m, which bought each roughly a quarter of that machine. Ongoing operations are funded through NCRIS and the partner collaboration. NCI is a public facility rather than a listed company, so there is no share listing or market valuation.

Its place in AI

NCI is Australia's national facility for high-performance computing and sits within the NCRIS network of research infrastructure. It hosts collections such as the Aboriginal and Torres Strait Islander Data Archive, the Australian Astronomy Optical Data Repository and the Australian National Geophysical Collection. High-resolution Indian Ocean sea floor data used in the search for Malaysia Airlines Flight 370 was generated on its systems. Its lineage runs back to ANU supercomputing from 1987.

Questions and criticism

The Vayu system drew about 605 kW at full load, and the plan was for all of that power to come from green sources. Cooling relies on free cooling and hot aisle containment to limit energy use. The facility depends on continued government and partner funding through NCRIS.

Questions people ask about National Computational Infrastructure

What is NCI Australia?

It is a national high-performance computing and data facility based at the Australian National University in Canberra. It is funded through the government's NCRIS programme and a collaboration of science agencies and universities, and it serves Australian researchers.

Who can use its computers?

Machine time goes to the funding partners, which include CSIRO, the Bureau of Meteorology, Geoscience Australia and several universities, and to researchers who win allocations through the National Computing Merit Allocation Scheme.

What is Gadi?

Gadi is NCI's main cluster, named from a Ngunnawal word meaning 'to search for'. It is rated at 9.26 PetaFLOP and uses Intel Xeon Scalable cores, some nodes with Nvidia V100 GPUs, connected over HDR InfiniBand.

How much data does it store?

NCI keeps 20 petabytes for fast input and output, 47 petabytes for large research files, and 50 petabytes on tape for long-term archives, along with hosted research collections in astronomy, geophysics and Indigenous data.

Where did NCI come from?

It traces back to the ANU Supercomputing Facility, active from 1987 to 1999, then the APAC National Facility commissioned in 2001. APAC began evolving into the present NCI collaboration in 2007.

Good reading and sources

Where this page comes from

The facts on this page come from the Wikipedia article "National Computational Infrastructure", written by its contributors and used under the CC BY-SA 4.0 licence, and from its Wikidata record (CC0). EveryNeocloud rewrote the text in its own words, with AI assistance, and compares every page with the article to make sure no passage is copied. It may be reused under the same CC BY-SA 4.0 licence with a link back.

It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.

Photo credits on this page

The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

The weekly digest

Every Monday: the provider of the day, new pages on EveryNeocloud, and the conferences and anniversaries of the week ahead.