HPC and supercomputing

Open source, fast, reliable, scalable and modular

Get a consistent and unified experience for HPC with solutions from Canonical across any type of cluster. On-premise, bare metal, private and public clouds.

Get in touch

Download our Guide to HPC ›

Advancing high-performance workloads

HPC is used across multiple sectors and industries

From energy to healthcare, we help industry leaders ease operations and break new ground with the power of compute.

Accelerate discovery in:

Agriculture
Automotive
Banking and finance
Construction
Defence and security

Electronics
Energy and utilities
Government
Healthcare and pharmaceuticals
Life sciences

Manufacturing and industry
Media and entertainment
Research and universities
Weather

Explore use cases

Solutions for HPC

Embrace the open source mandate in HPC with Canonical

Cluster management at scale

MAAS: super fast cluster deployment and delivery

The ultimate experience for bare metal resource discovery, management, delivery and observability. Deploy HPC and supercomputing clusters with ease. Test and validate clusters with commissioning scripts.

Read more about cluster architecture ›

Operating system for computational workloads

Ubuntu is the distro for HPC, HPDA and AI/ML

Ubuntu is the ultimate distribution for HPC. Recent kernel , LTS releases with 5-year support, maintenance and bug fixes, extendable to 10 years. An extensive list of packages available across its repositories, giving you a unified source for installation. Available for local installation or public clouds. Perfect for long-running environments.

Software and cluster automation

Juju and Charms: deploy, maintain and operate, on autopilot

Simplify the delivery of HPC clusters. Charmed HPC software solutions could be just the thing you need to ease cluster operation and monitoring.

Juju charms help you simplify operations by assisting in the installation, maintenance and integration of applications.

Container-based resource scheduling

Charmed Kubernetes: a unified experience across substrates

Manage your cloud-native computational workloads with ease using Charmed Kubernetes.

Management for virtualised computational infrastructure

OpenStack: the reference architecture for computational workloads

Get agile management for your computational resources alongside an extended service portfolio for VDI and auxiliary nodes. Everything to support your computational workload.

Scalable storage for objects and files

Ceph: a scale out solution

CephFS delivers a scale-out parallel file system for workloads requiring POSIX support. Deliver archival tiers, where supported, with object storage.

Scania, a major manufacturer of commercial vehicles, relies on SLURM and a Canonical stack to power up and scale their HPC clusters. Using MAAS and Juju, they are well-positioned to meet peaks in demand with confidence.

Watch the video ›

The Wellcome Sanger Institute is a global centre of excellence for genomic research. As capacity needs grew from 3 PB to over 20 PB, the Institute turned to Canonical for high-level Ceph support, providing direct access to Canonical's engineering teams to resolve the more complex issues.

Read the story ›

Performance through partnerships

Canonical has a strong network of technical alliances, driving platform compatibility on Ubuntu. Close partnerships make sure Ubuntu runs optimally on the CPU architecture and works well with GPUs, FPGA, accelerators and high speed interconnects.

Learn more about partnerships

Driver availability and NGC toolchain support

Get optimal performance with NVIDIA Infiniband and GPUs through driver availability on Ubuntu. Take advantage of NGC containers for your workloads to accelerate innovation.