Deploy virtual machines, bare metal servers, and GPU instances with enterprise-grade performance and flexibility across any architecture.
Experience the power of rapid provisioning, multi-architecture support, GPU capabilities, network offloading, and more. Our Compute Service is designed to deliver performance, flexibility, and control.
Rapidly create and manage virtual machines, providing a flexible and scalable environment to meet your specific needs and workloads.
Efficiently manage physical resources with the ability to provision baremetal resources, providing high performance for resource-intensive tasks.
Accommodate diverse workloads with compatibility for multiple architectures including x86 and ARM, offering versatility for computing needs.
Optimize resource-intensive tasks with dedicated GPU instances. Enable a variety of use cases with vGPU support.
Expedite data processing with SR-IOV, DPDPK, and ASAP2, enhancing the performance of your cloud.
Tailor cloud resources to fit your specific needs, offering flexibility and control over the computing environment.
Maintain continuous operation with the ability to move running VMs without downtime, offering flexibility in resource management.
Stay ahead with services offering native support for IPv6, ensuring your cloud infrastructure aligns with modern network standards.
Operate a virtual machine within another, enhancing resource utilization and offering greater flexibility.
All of our public cloud services are available across our strategically located datacenters, covering the United States, Canada and Europe, with strong local connectivity to minimize latency to your infrastructure.
No complicated payment structure. You’re simply charged for your usage, down to the minute. If your server continues to run for over a month, you get our discounted monthly rate applied automatically.
We’ve built our cloud from the ground using open source, with no local forks. Everything we deploy is contributed upstream. We also sponsor open source foundations & many projects (such as curl!) for CI resources.
Starter servers are low-cost, general-purpose instance types that provide a small burstable portion of virtual cores and 2 gigabytes of memory per virtual core.
Optimized for: Burstable workloads such as low traffic web servers (blogs, forums, CMS), small databases, development or test systems, code repositories, and prototype systems.
Standard servers provide cost-effective & class-leading computing with dedicated physical processor threads and 4 gigabytes of memory per core.
Optimized for: High-performance CPU-bound workloads such as the production of web applications, e-commerce websites, enterprise applications, or CI/CD build systems.
GPU cloud servers provide you with production-level high-performance computing. These servers deliver powerful virtualized NVIDIA K80 GPUs.
Optimized for: Heavy workloads such as high-performance computing (HPC), artificial intelligence (AI), or machine learning (ML), run by researchers, data scientists, and developers.
Only pay for the block storage you need by adding volumes to your instances based on your app’s requirement.
Easily access our quick, scalable, and secure object storage solution to confidently store any data you need.
Distribute workloads across multiple computing resources to eliminate interruptions in a quick and easy manner.
Effortlessly enable IP blocks and configure multiple IP addresses to your cloud server while allowing for failover.
Spin up an instance with a choice list of operating systems. Select the OS that fulfills your requirements and get started quickly.
Upload your own images to launch VMs and get more flexibility by uploading your own customized operating system, external snapshot image in RAW format, or virtual appliance.
You can combine firewall rules (based on source & destination IP addresses, ports and protocols) into groups (also known as “security groups” in cloud jargon) and apply those groups into your servers. If you make any changes to these rules, your changes will propagate to all of your servers instantly. Also, the denied traffic won’t even show up to your servers’ network interfaces.
Take snapshots of on-demand images of VMs and volumes, enable backups for automatic weekly VM images, and upload custom images to create VMs with other operating systems or pre-packaged libraries. All these aspects will help you roll back to an original server configuration if need be.
We understand how critical downtime is. That’s why we offer a 99.99% uptime SLA and credit you for any downtime experienced.
DNS records can easily be created and customized. DNS provides you with a REST API along with integrated identity authentication.
Our data centers in Montreal, Santa Clara, and Amsterdam make it possible to span your cloud environment across multiple regions.
The trailblazing speed of network connectivity will outlast all performance expectations and project goals.
Use the most recent version of internet protocol to map your virtual machines in the cloud.
Enable each VM within a given hypervisor to communicate privately with one another.
Restore an existing snapshot to a different cluster with the option to share it with other users.
Automate package installation during initial setup through personalized user scripts.
Launch multiple servers within a few seconds from our Cloud Console or from the command line interface via API.
Dynamically scale up or scale down any of your compute resources without any complex steps.
Take advantage of increased performance with enterprise-grade hardware and NVMe and SSD storage options.
Get increased security and network performance through our powerful KVM hypervisors.
🎊Atmosphere is an open source project built by us.
Atmosphere stands out as a fully certified and conformant open-source platform. With both OpenStack powered certification and certified Kubernetes, it has passed all conformance tests. This guarantees a robust, high-performance, and reliable solution for your cloud needs.
Build and operate production-grade private and public clouds with OpenStack. Our certified engineers provide end-to-end professional services from architecture design to ongoing operations.
Experiencing a critical incident? Indicate the urgency in your message for immediate 24x7 emergency support from our expert team.
Let us help you with your OpenStack needs
Insights, updates, and stories from our team
Learn how to build zero-trust cloud infrastructure using OpenStack and Atmosphere. Discover security-by-default practices for enterprise private cloud in 2026.
Predictable cloud costs drive better planning. Learn why OpenStack-powered infrastructure delivers the cost visibility hyperscalers can't match.
Kubernetes is everywhere in 2026—but where you run it matters. Explore infrastructure trade-offs between hyperscalers and OpenStack.
