{
  "id": "cuda/nccl-ib-device-not-found",
  "signature": "RuntimeError: NCCL error: unhandled system error, NCCL version 2.18.5, error: IB device not found",
  "signature_zh": "运行时错误：NCCL 错误：未处理的系统错误，NCCL 版本 2.18.5，错误：未找到 IB 设备",
  "regex": "IB device not found",
  "domain": "cuda",
  "category": "network_error",
  "subcategory": null,
  "root_cause": "NCCL (NVIDIA Collective Communications Library) attempts to use InfiniBand (IB) for inter-GPU communication but cannot find any IB device, either because the IB driver is not loaded or the hardware is not present.",
  "root_cause_type": "generic",
  "root_cause_zh": "NCCL（NVIDIA 集体通信库）尝试使用 InfiniBand (IB) 进行 GPU 间通信，但找不到任何 IB 设备，可能是因为 IB 驱动程序未加载或硬件不存在。",
  "versions": [
    {
      "version": "NCCL 2.18.5",
      "introduced": null,
      "deprecated": null,
      "removed": null,
      "behavior_change": null,
      "status": "active"
    },
    {
      "version": "NCCL 2.19.3",
      "introduced": null,
      "deprecated": null,
      "removed": null,
      "behavior_change": null,
      "status": "active"
    },
    {
      "version": "CUDA 11.8",
      "introduced": null,
      "deprecated": null,
      "removed": null,
      "behavior_change": null,
      "status": "active"
    },
    {
      "version": "CUDA 12.1",
      "introduced": null,
      "deprecated": null,
      "removed": null,
      "behavior_change": null,
      "status": "active"
    }
  ],
  "os_specific": {},
  "dead_ends": [
    {
      "action": "",
      "why_fails": "NCCL will still probe for IB devices; if hardware is absent, the error persists regardless of NCCL version.",
      "fail_rate": 0.9,
      "condition": "",
      "sources": []
    },
    {
      "action": "",
      "why_fails": "The error is fatal; NCCL will abort the communication initialization, not just log a warning.",
      "fail_rate": 0.95,
      "condition": "",
      "sources": []
    }
  ],
  "workarounds": [
    {
      "action": "Set the environment variable NCCL_IB_DISABLE=1 to force NCCL to use TCP/IP instead of InfiniBand for inter-node communication.",
      "success_rate": 0.85,
      "how": "Set the environment variable NCCL_IB_DISABLE=1 to force NCCL to use TCP/IP instead of InfiniBand for inter-node communication.",
      "condition": "",
      "sources": []
    },
    {
      "action": "If IB hardware is available but driver not loaded, load the InfiniBand kernel module: `modprobe ib_core` and verify with `ibstat`.",
      "success_rate": 0.75,
      "how": "If IB hardware is available but driver not loaded, load the InfiniBand kernel module: `modprobe ib_core` and verify with `ibstat`.",
      "condition": "",
      "sources": []
    }
  ],
  "workarounds_zh": [
    "Set the environment variable NCCL_IB_DISABLE=1 to force NCCL to use TCP/IP instead of InfiniBand for inter-node communication.",
    "If IB hardware is available but driver not loaded, load the InfiniBand kernel module: `modprobe ib_core` and verify with `ibstat`."
  ],
  "transition_graph": {
    "leads_to": [],
    "preceded_by": [],
    "frequently_confused_with": []
  },
  "official_doc_url": "https://docs.nvidia.com/deeplearning/nccl/user-guide/docs/env.html#nccl-ib-disable",
  "official_doc_section": null,
  "error_code": null,
  "verification_tier": "ai_generated",
  "confidence": 0.86,
  "fix_success_rate": 0.8,
  "resolvable": "partial",
  "first_seen": "2023-08-01",
  "last_confirmed": "2024-06-01",
  "last_updated": "2024-06-01",
  "evidence_count": 1,
  "tags": [],
  "locale": "en",
  "aliases": []
}