Data In Situ Computation

Currently, data is being underused, as the restrictions and regulations around data sharing are stifling innovation.

NVM Compute Functionalities help you turn datasets into Web3-ready assets through its remote, or federated, computing capabilities. The result: you can develop new business models and ecosystems around data to create innovative new products and value streams.

Compute Utility 1 - Federated Computation

What is it?
The ability to deploy algorithms and compute on disparate, or “federated” datasets. In other words, move compute to data, not the other way around. Support for multiple Federated Learning frameworks, including Google’s FL framework.

Why use this?
Federated Learning enables algorithmic training and model deployment to decentralized assets without the need to consolidate those assets into centralized repositories. The result is that users can now utilize machine learning and AI remotely without having to move data.

Read about how Federated Computation can be used or dive into the Specs of our Remote Compute Service .

Compute Utility 2 - ML & AI Model Royalties

What is it?
A way to tokenize algorithmic models and add Royalty utility to it.

Why use this?
This allows data scientists to create new revenue models around their assets.

Discover our SDK for more.

Compute Utility 3 - AI & Machine Learning Industrialization

What is it?
A way to execute Distributed Computation or Massive Parallel Processing (MPP) using analytics frameworks (i.e Spark, Flink).

Why use this?
These common frameworks are extremely powerful, enabling the processing of exabytes of data on cheap, commodity hardware. In addition, certain frameworks support the training and deployment of AI.

Discover our SDK for more.