IBM Cloud Docs
Public virtual servers

Public virtual servers

The public IBM Cloud® Virtual Servers offerings are IBM-managed, multi-tenant virtual server deployments. They give you rapid scalability and higher-cost effectiveness with pre-defined sizes that meet all business requirements to get you up and running quickly.

Newer version available! Try our Virtual Servers for VPC. For more information, see Virtual Private Cloud.

Public virtual servers have many advantages. The following are some examples.

  • Global availability The public virtual server offering is available in data centers across the globe.
  • Configuration choices, rapid deployment, and scalability The public virtual server offering gives you small or large virtual server options to meet various workload requirements.

Through extra configuration, network performance of virtual servers can be improved as compared to using the default settings. For more information, see Configuring virtual server settings for improved network performance. However, the amount of allowable network traffic for public virtual servers, which encompasses virtual server instance SAN and other network-attached storage, has no guarantee. If network traffic of a virtual server instance has a significant, negative impact on other virtual servers, that instance might restart on a new host, or in extreme cases, become disabled. This negative impact is often observed as traffic levels approach 20,000 - 30,000 packets per second (PPS). For guaranteed networking, use of the Dedicated Virtual Server offering is recommended. For more information, see the single-tenant environment, Dedicated virtual server offering.

The following families of public instances are available for this offering.

Public virtual server family selections
Families Description
Balanced Best for common cloud workloads that require a balance of performance and scalability. Uses network-attached storage.
Balanced local storage Best for large database workloads that require high I/O performance with low latency.
Variable compute Best for workloads that don’t require steady, high-CPU performance.
Compute Best for moderate to high web traffic workloads.
Memory Best for memory caching and real-time analytics workloads.
GPU Best for high-performance workloads.

Next steps

After you review and decide upon your virtual server profile, it's time to provision your public virtual server. To get started, see Provisioning public instances.