This describes a proposed reference architecture. It is not a shipping product and no vendor partnership is claimed or implied. Interfaces and standards conformance are specification intent; quantities are design targets.
Declarative infrastructure
Cluster state is described declaratively and applied repeatably, so a rebuild is a procedure rather than an archaeology exercise.
Workload manager
Job scheduling, queueing, and fair-share policy across the cluster, extended with the plant-aware inputs described above.
Container and service layer
Service orchestration for inference and platform workloads, with quota and isolation between tenants.
Unified observability
Node, fabric, thermal, and power telemetry in one pipeline shared with the plant supervisory layer rather than a separate silo.
Component selection follows open standards; specific products are procurement decisions. The differentiator is the plant tie-in, not the individual components.
Discuss the details.
Specifications evolve as the design matures. Contact us for current status and integration requirements.
CONTACT TIMOTHY → Back to Run by Plant Intelligence →