{
  "slug": "clio",
  "term": "Clio",
  "aliases": [
    "Clio disaggregated memory"
  ],
  "definition": "Clio is a research disaggregated-memory system that combines a client library with hardware memory nodes for address translation, protection and remote access.",
  "note": "Architecture: CBoard separates a hardware data path from software control on ARM. CLib manages client-side request ordering. The design is not ordinary one-sided RDMA on a memory server. [paper]\n\nAddress trade-off: Hash-bucket overflow is avoided when allocating virtual addresses, moving retries to allocation rather than lookup. A requested fixed address is not always available. [paper]\n\nRetry boundary: Memory nodes retain recent write identifiers and atomic results. The described buffer covers two retries; exhaustion is reported to the application, not an unlimited exactly-once guarantee. [paper]\n\nFailure boundary: Request retransmission is not recovery from a failed memory node. The paper leaves built-in failure handling and replicated-write primitives for future work. [paper]\n\nMeasurement boundary: Figure 8 uses 1 KB requests and reaches a 9.4 Gbps end-to-end testbed ceiling. Figure 9 exceeds 110 Gbps for large requests using on-board traffic generation that bypasses the 10 Gbps port. These are different measurement paths. Figure 6 labels ASIC latency as projected. [paper]\n\nReproduction scope: The guide selects microbenchmarks, not every application or on-board experiment. Switched and direct-connected setups differ; the split cable adds latency. [artifact]\n\nBuild boundary: The guide targets Vivado/HLS 2019.1.3, Scala, ARM cross-compilation and host software. Other Vivado versions may need IP changes. [build]\n\nAPI naming: The header declares legomem_alloc, legomem_read, legomem_write_sync and legomem_write_async, not the paper's ralloc/rread/rwrite names. Declarations alone do not prove implementation guarantees. [api]",
  "status": "research",
  "advantages": [],
  "limitations": [],
  "related": [
    "infiniswap",
    "rdma-storage"
  ],
  "supersedes": [],
  "supersededBy": [],
  "firstSeen": null,
  "lastSeen": null,
  "updated": "2026-11-15",
  "citations": [
    {
      "id": "paper",
      "title": "Guo et al. Clio: A Hardware-Software Co-Designed Disaggregated Memory System. ASPLOS 2022, pp. 417-433; sections 4.2, 4.5, 5, 7, 9.",
      "url": "https://cseweb.ucsd.edu/~yiying/clio-asplos22.pdf",
      "accessed": "2026-11-15"
    },
    {
      "id": "artifact",
      "title": "Clio authors: artifact evaluation scope, selected microbenchmarks and two testbed configurations, pinned revision 7908c3a",
      "url": "https://github.com/WukLab/Clio/blob/7908c3a022fd356cd54616630fcddf59f7f3fd95/Documentation/asplos-ae.md",
      "accessed": "2026-11-15"
    },
    {
      "id": "build",
      "title": "Clio authors: hardware, ARM and host compilation prerequisites, pinned revision 7908c3a",
      "url": "https://github.com/WukLab/Clio/blob/7908c3a022fd356cd54616630fcddf59f7f3fd95/Documentation/compile.md",
      "accessed": "2026-11-15"
    },
    {
      "id": "api",
      "title": "Clio authors: public host API declarations, pinned revision 7908c3a",
      "url": "https://github.com/WukLab/Clio/blob/7908c3a022fd356cd54616630fcddf59f7f3fd95/host/api.h",
      "accessed": "2026-11-15"
    }
  ],
  "provenance": [
    "Primary-source synthesis of the final paper and pinned author documentation/header. No FPGA, RDMA, Clio runtime or benchmark was executed by Hesela.",
    "Throughput numbers are author-reported, not Hesela measurements. No confidence intervals or general production-speed advantage are inferred.",
    "The code revision is a reproducible inspection target, not proof that it produced every result in the paper."
  ]
}
