We are developing 3 interrelated platforms that will make it easy to create Big Control apps.
A framework that simplifies the creation of Big Control applications. Provides control-specific primitives for tasks such as data ingestion and fusion, declarative planning, and deep reinforcement learning. Leverages the facilities of the Granular Computing Platform and Self-Programming Networks.
GRANULAR COMPUTING PLATFORM
Enables efficient execution in a datacenter of a very large number of very short-lived tasks (from a few milliseconds down to a few tens of microseconds). Used by swarm applications to handle millions of device notifications per second, and to subdivide large calculations into tiny pieces so they can complete in real time. Depends on self-programming networks for communication, both with devices and within the datacenter.
SELF-PROGRAMMING NETWORKS (SPN)
Given a few high-level goals, generates the precise control policy for a network automatically by sensing load and applying machine learning algorithms. Delivers deterministically low latency to Big Control for drastically varying workloads.