Ontology of Architectural Regimes in Computing: Growth Fronts, Closure, and Architectural Disposal Beyond Evolutionary Narratives

This work develops an architectural ontology of computing that rejects evolutionary and progress-oriented narratives of technological change. Rather than treating technical objects as entities that develop or evolve, the paper argues that computing history is structured by a sequence of externally governed architectural regimes. Within this framework, architectures exist as unified entities only while a regime of structural coherence can be maintained. When unresolved structural tensions exceed the integrative capacity of a regime, architectural unity is lost and replacement occurs. Change is therefore discontinuous and ontological rather than incremental or adaptive. The paper introduces a regime-level vocabulary for computing architectures, centered on three structural operators: growth fronts (zones of architectural indeterminacy), closure points (fixation of architectural invariants enabling reproduction), and odd remainders (non-compensable structural tensions that drive degradation and regime termination).Architectural disposal is formalized as an ontological event: the loss of a regime of unity, distinct from market obsolescence or lifecycle decline. Using this framework, the history of computing is reconstructed as a sequence of regimes, including pre–von Neumann plurality, the von Neumann closure and its long reproduction phase, the transition to parallel and multi-core architectures under integration limits, and contemporary cloud and distributed systems as a partially closed regime. Persistent tensionsin distributed systems, including those formalized by CAP, are interpreted as diagnostic indicators of unresolved architectural closure rather than as solvable trade-offs. The contribution of this work is not a new chronology of computing or a theory of innovation, but a structurally grounded language for distinguishing architectural existence from dynamic persistence, reproduction from development, and replacement from improvement. The framework supports diagnostic analysis of contemporary computing without resorting to technologicalfuturism or evolutionary metaphors.