Choose hourly billikng and pay only for what you use.

Choose monthly billing for consistent pricing and budgeting.

Help reduce costs and stop reprovisioning servers when you need resources again. Suspend billing is a fast, cost-efficient answer to stopping and reclaiming capacity for hourly virtual servers.

Help reduce costs by running specific workloads, such as non-production or dev/test, on transient servers with on-the-spot pricing. These servers are de-provisioned on a first-on, first-off basis.

Public, multi-tenant servers come in a variety of pre-configured profiles to better suit your dynamic workload requirements. Profiles include: Compute, Variable Compute, Memory, GPU, Balanced and Balanced Local Storage.

Dedicated, single-tenant servers offer added control and isolation for sensitive workload requirements. You choose dedicated hosts or dedicated instances. Dedicated hosts help with control over workload placement and dedicated instances offer single-tenant isolation.

Transient servers—also known as “spot pricing” servers—are de-provisioned on a first-on, first-off basis; are offered at a significantly lower price; and are ideal for dev/test applications and other non-production workloads that don’t require constant uptime.

Reserve up to 20 virtual servers in advance, in the data center of your choice. Choose 1-year or 3-year contract terms. Reserved virtual servers are ideal for sustained workloads that require enterprise-level budgeting outlooks.

Single-tenant dedicated hosts help control workload placement and support network throughputs of up to 20 Gbps. View core, RAM and local storage consumption for maximum control of workload management.

Ideal for CPU-intensive workloads, such as front-end processing and batch processing

Ideal for low-usage and intermittent workloads that don’t require consistent CPU-capacity performance

Ideal for memory-intensive workloads, like in-memory analytics and medium-to-large caching

Ideal for high-performance workloads requiring more compute density, such as AI, intense graphics and data applications, and deep learning

Ideal for common cloud workloads requiring a balance of performance and scale, such as moderate web traffic

Ideal for large database workloads requiring low latency and high I/O performance. Select from SSD and HDD storage options.

Deploy virtual server workloads in over 60 IBM Cloud Data Centers across 6 regions and 19 availability zones globally. Built for local access, low latency and certified security, IBM Cloud offers a range of choices about where and how your data and workloads run. The availability zone design can make applications and databases highly available, fault-tolerant and scalable.

All inbound and outbound bandwidth inside the IBM Cloud global, private network is unlimited and cost-free for your virtual server use.

Public, outbound bandwidth is charged on a tiered basis, with a set allotment for each month of your virtual server use.

IBM Cloud Virtual Servers on the IBM Cloud VPC network deliver fast provisioning compute capacity with the highest network speeds and most secure, software-defined networking resources available on the IBM Cloud.