An Off-The-Shelf Private Cloud
Compute Groups virtualize your enterprise hardware into composable computing resources. Depending on how much hardware you own, one could add tens of thousands of core hours to your reserve without buying a single server.
A Compute Group can exist across the entire organization, or it can be partitioned into different pools owned by different business units. These units can easily account for their consumption, and share their excess resources with others.
Compute Groups are just $20/core year or $100/GPU year, making them the cheapest private cloud option by an order of magnitude.
Recapture every dollar of idle resources from your hardware without having to do any container management or orchestration
Manage your enterprise resources with the same simple Compute API, exercising granular control over who can use what resources and when
Greater on-premise capacity means there is less of a need for your data to go to the cloud, paired with best-in-class user authentication
Unlocking Hidden Potential
Compute Groups connect ordinary desktops, laptops, and even IoT devices together with your main server resources. These create powerful networks, which can then be connected or adjoined together to make an infinitely expandable private cloud.
Just like a smart electrical grid, excess computing power from your private network can be sold back to recover your own costs.. It is even possible to profit by selling more than you use back to the Distributed Computer.
Unlike similar solutions which rely on containerization, each node requires almost no overhead to perform powerful compute operations. These do not interfere with regular use of the device at all.
Compute Groups do not send any data outside your network firewall.. They automatically schedule workloads to take advantage of the appropriate infrastructure, so your most capable hardware is freed to deal with the most demanding tasks.
What Are Some Use Cases?
Compute Groups excel in cases where data cannot leave the premises, more server hardware is expensive or difficult to procure, when IT staff are short on time, or when workloads regularly need to burst to the public cloud.About Hybrid Solutions →
Higher Education Grids
Thousands of cores on a university campus can be unlocked and combined into a hyperscale compute platform, empowering researchers and students.
Compute Groups are an ideal pre-production environment to locally test applications that will have to be scaled in an off-premise cloud.
Different business units can combine or share their resources, and applications can be shared across the organization that are powered by its own infrastructure.
Capturing Off-Peak Resources
A multinational organization can shift workloads around time zones to capture idle capacity in its servers.
Keeping Workloads Compliant
Compute Groups offer a way to keep algorithms and data completely within your firewall, while significantly increasing your on-premises throughput.