System/360 Storage Hierarchy and Cache Model
Scope
Cache behavior is implementation-dependent and must not be assumed to be a universal System/360 architectural feature. Some models used different storage technologies and performance characteristics.
Modeling levels
- Architectural: storage results, protection, interruption, and ordering behavior.
- Performance abstraction: configurable access latency based on documented machine data.
- Implementation detail: cache or buffer behavior only when supported by the selected model's documentation.
Main storage
Represent installed main storage, address ranges, access restrictions, and wait states in the machine profile. Keep these separate from optional performance abstractions.
Evidence
Label storage latency as documented, derived, estimated, or unspecified. Do not assign modern cache structures to historical models without evidence.