OneCompute
| Release Notes | Select another OneCompute version |
|---|---|
Links to OneCompute Documentation
OneCompute Examples
Example code for using OneCompute can be found on GitHub:
Where to find OneCompute
Using NuGet
- NuGet.org Note: NuGet only contains the libraries needed for using the OneCompute Platform
- OneFoundation feed on Azure DevOps (DNV internal feed): https://pkgs.dev.azure.com/dnvgl-one/_packaging/One/nuget/v3/index.json
About OneCompute
OneCompute aims at answering the following questions:
- How can users and developers of computational applications benefit from running their computational workflows in the cloud?
- How can computational workflows be delivered to the end user without having to set up and manage any application specific services or infrastructure?
- How can applications, or even end users, create and execute complex custom computational workflows?
OneCompute comes in two shapes, depending on the constraints of the application:
- OneCompute Platform is an end-to-end cloud platform for hosting computational applications. This is the recommended option for any application that is not restricted from using public cloud resources, because it will immediately benefit from the scalability of the cloud as well as the functional and technical features brought by the OneCompute Platform itself that make application development, deployment and operation simpler and more cost efficient.
- OneCompute Libraries is a general-purpose and platform independent framework for defining computational workflows, referred to as jobs, and components for orchestrating them. As a framework, OneCompute Libraries does not prescribe how to deploy and execute computational services, leaving that as a decision to be made by the application. Both on-premise and private cloud deployments can be supported by using this option.
Notice that OneCompute Platform builds on OneCompute Libraries, and they share the basic APIs and programming model.
OneCompute targets application scenarios where it makes sense to offload computational work to a dedicated service instead of orchestrating it as part of the application itself. Typically, these are cases where the amount of computational work to perform is too time consuming to execute without offloading it to an external computational service that has more resources available and is capable of conducting the work more efficiently. OneCompute Platform offers end-to-end services to this effect. It provides a deployment option for computational applications that does not require any application specific services to be set up.