{
  "slug": "miss-ratio-curve",
  "term": "Miss-ratio curve",
  "aliases": [
    "MRC",
    "cache miss curve"
  ],
  "definition": "A miss-ratio curve relates cache capacity to the fraction of requests that miss under a specified workload, policy and initial state.",
  "note": "For equal-size demand-admitted LRU, count cold accesses plus reuses whose zero-based distinct-entry distance is at least the capacity, then divide by all requests. A one-based stack-distance formulation instead counts distances exceeding capacity. [counter-stacks]\n\nIn the cold-start trace 0,1,2,0,0,2,1, capacities 1, 2 and 3 give 6/7, 5/7 and 3/7 misses. These are constructed counts, not device-speed predictions. [hesela]\n\nA trace starting mid-workload does not reveal the initial cache contents. Preserve the observation layer, history and time window; a whole-run aggregate can hide changing locality. [counter-stacks]",
  "status": "research",
  "advantages": [],
  "limitations": [],
  "related": [
    "reuse-distance",
    "lru-replacement",
    "cache-inclusion-property",
    "belady-anomaly"
  ],
  "supersedes": [],
  "supersededBy": [],
  "firstSeen": null,
  "lastSeen": null,
  "updated": "2026-10-19",
  "citations": [
    {
      "id": "counter-stacks",
      "title": "Wires et al. Characterizing Storage Workloads with Counter Stacks. OSDI 2014, sections 2 and 3.2.",
      "url": "https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-wires.pdf",
      "accessed": "2026-10-19"
    },
    {
      "id": "hesela",
      "title": "Hesela. LRU sizing from distinct-entry reuse distances; constructed examples.",
      "url": "https://hesela.com/analyses/ssd-mapping-cache-hmb/#reuse-distance",
      "accessed": "2026-10-19"
    }
  ]
}