We are ClusterVision
HPC Cluster Solutions | ClusterVision
ClusterVision is a European hardware and software company that specialises in providing end-to-end High-Performance Computing (HPC) and Artificial Intelligence (AI) solutions, headquartered in the Netherlands.
Alex Ninaber and Dr. Paul Calleja started experimenting with the concept of Beowulf computing at Birkbeck College in 1998, as part of the research group of Dame Julia Goodfellow — an approach introduced amongst others by Thomas Sterling and Donald Becker at NASA, which has since revolutionised the high-performance computing industry.
Starting from a division of Compusys Ltd in 1999, together with Dr. Paul Calleja (Director of Research Computing at Cambridge University), Ninaber and Calleja were among the first in Europe to build HPC clusters using commodity hardware and open-source software.
The company was co-founded in 2002 by Alex Ninaber and Matthijs van Leeuwen.
Our History
Hailing from a computer science background and with deep prowess in HPC applications, the founders started the venture to meet the surging need for professional partners adept in the domain of HPC computations and applications.
ClusterVision built one of its first supercomputers in 2003, for Manchester University in the UK. The Linux-based supercomputer comprised 70 processors, based on the then-new 64-bit AMD Opteron processors. Its first use was by the Faculty of Chemistry at Manchester University, to carry out simulations for molecular systems and chemical reactions.
ClusterVision’s first cluster management software was developed in 2004, and made commercially available in 2008 under the name ClusterVisionOS v4. In 2009, the company spun this off into a separate entity named Bright Computing; the software was renamed Bright Cluster Manager, and the spun-off company is now part of NVIDIA.
Acquisition by Taurus Group
In February 2019, ClusterVision was acquired by the Taurus Group, a large computer hardware distribution company dealing in components for desktop PCs and servers.
With the financial strength of Taurus and several key ClusterVision employees on board, ClusterVision now trades as ClusterVision Solutions, while continuing to operate under the ClusterVision name. The development of its in-house cluster management software, TrinityX, also gained a financial boost.
The company now sees significant growth ahead for its business in the field of supercomputing, as data volumes continue to grow at an accelerated rate.
Our Customers
ClusterVision has clients on all major continents, with a strong presence in the EMEA region. The company has clients in more than 10 vertical markets, including automotive, healthcare, education, oil and gas, and life sciences.
During our existence, ClusterVision has served more than 700 clients and installed more than 2,000 projects.
Milestones & Accomplishments
The Top500, initiated in 1993, ranks the world’s most powerful supercomputers using the Linpack Benchmark, which evaluates their capacity to solve linear equations. Rankings are updated twice a year, in June and November.
ClusterVision’s notable installations include partnering with Dell in 2006 to secure 20th place for Europe’s fastest x86 HPC cluster, at Cambridge University. In 2010, the supercomputer Millipede ranked 414th at the University of Groningen (RUG), and Tetralith ranked 75th in both 2015 and 2018, with a Linpack performance of 2,969.19 TFlop/s.
A particular highlight was the 2014 build of the Vienna Scientific Cluster’s oil-cooled VSC-3 at TU Wien — the world’s first fully oil-submerged supercomputer — which earned ClusterVision 85th place in the Top500.
Products and Services
ClusterVision is well known for combining ground-breaking hardware with state-of-the-art software to deploy, manage and support HPC clusters, backed by reliable support solutions.
Our hardware is engineered to deploy turnkey or tailor-made clusters for our customers’ complex HPC needs, integrating strong compute, storage and GPUs, backed by services and remote admin support.
ClusterVision provides an end-to-end service, right from hardware installation through to keeping the cluster running — even from a remote location — and specialises in fully equipped, technologically advanced software solutions that meet all of our clients’ HPC needs.
In 2013, ClusterVision started developing its new cluster management software environment, TrinityX. Unlike the Bright Cluster Manager, it was decided to fully open-source the software under the GPL 2.0 licence.
Initially based on OpenStack, the team decided in 2016 to abandon that path and start from scratch, using a torrent-based approach for Exascale scaling. The code was released for production in 2017.
TrinityX has been meticulously developed by ClusterVision’s in-house technical team, and delivers strong performance even while managing the most complex HPC computations for our clients.
Memberships, Frameworks & Certifications
ClusterVision is certified by BSI for ISO 9001 (Quality Management), ISO 14001 (Environmental Management), and ISO 27001 (Information Security Management). We have an excellent track record on ISO certifications, holding our internal management and reporting processes to the highest standards.