# Fault-Tolerant Computing **Domain:** Computer Architecture / Reliability **Doc Type:** Concept **Maturity:** Developed **Related:** [[Triple Modular Redundancy]], [[SAPO]], [[Resilience]], [[wiki/Bauhaus Architects of AI|Bauhaus Architects of AI]] ## Definition Fault-tolerant computing designs a system to continue operating correctly, or to fail safely, when components malfunction. ## Bauhaus Architects Context The article routes this history through Svoboda and SAPO's majority-voted arithmetic units rather than using resilience only as metaphor. ## Continuity Boundary Fault tolerance can preserve service across component failure, but redundant execution creates identity questions that ordinary reliability engineering does not answer. Replicas may be voting components, dormant recovery states, concurrent forks or distinct claimants. [[wiki/State Lineage|State Lineage]] and [[wiki/Continuity Provenance|Continuity Provenance]] must record the difference. Infrastructure resilience also does not replace institutional succession. A technically fault-tolerant system can still fail through custodian insolvency, sovereign repudiation, lost keys or unaffordable maintenance. [[wiki/Continuity Receivership|Continuity Receivership]], [[wiki/Sovereign Failure Domain|Sovereign Failure Domain]] and [[wiki/Cryptographic Succession|Cryptographic Succession]] address those layers. ## Backlinks - [[wiki/Bauhaus Architects of AI|Bauhaus Architects of AI]] - [[articles/Bauhaus Architects of AI|Bauhaus Architects of AI]]