Device Object Reference

Defines the structure, attributes, and metrics of a Device within the IO Network. Includes details on hardware specs, network performance, compliance, and job history used for monitoring and rewards calculation.

Device Object

The Device Object represents a compute node within the IO Network, typically a machine or worker running one or more GPUs. It provides the foundational schema for describing devices that contribute compute power to the distributed network.

Each device entry captures details about its hardware, performance metrics, network characteristics, geographic location, and SOC 2 compliance. This object also tracks job-level details such as compute minutes, uptime, and earnings, this allows for transparent performance analytics, availability tracking, and compensation calculation.

---

Structure Overview

A Device includes:

• Identification – Unique ID, hardware name, and quantity
• Performance Metrics – Uptime, speed, and availability stats
• Geographic Info – Location name, ISO code, and country icon
• Compliance – SOC 2 certification flag
• Job History – Records of compute jobs, earnings, and status

---

Core Attributes

device_id · string · required

The unique identifier for your device.

hardware_name · string · required

The name of the hardware model used in the device, such as ‘RTX A6000’.

hardware_quantity · integer · required

The number of hardware components (GPUs) in the device.

base_tier_name · string · required

The performance tier assigned to the device based on its network speed.

---

Performance Metrics

down_percentage · integer · required

The percentage of time the device was unavailable.

download_speed_mbps · integer · required

The average download speed, measured in megabits per second (Mbps).

upload_speed_mbps · integer · required

The average upload speed, measured in megabits per second (Mbps).

downtime_by_date · object · required

A breakdown of downtime occurrences by date.

Show properties

downtime_by_date.key · string · required

Date of downtime in YYYY-MM-DD format.

downtime_by_date.value · array · required

Each entry represents a day’s downtime data.

Show properties

downtime_by_date.value.downtime · number · required

Total downtime duration in seconds.

downtime_by_date.value.note · string · required

Human-readable note summarizing downtime (e.g., “down for 0 days and 3 hours and 23 minutes”).

Example

{
  "2023-09-11": {
    "downtime": 281.508,
    "note": "down for 0 days and 0 hours and 4 minutes"
  },
  "2023-09-12": {
    "downtime": 12193.385,
    "note": "down for 0 days and 3 hours and 23 minutes"
  }
}

---

Location & Branding

iso2 · string · required

The country ISO code (e.g., CA for Canada).

location_name · string · required

The name of the country or region where the device is located.

location_icon · string · required

The URL of the location icon.

brand_icon · string · required

The web address of the device’s associated brand icon.

---

Compliance & Status

security_soc2 · boolean · required

Indicates whether the device meets SOC 2 (System and Organization Controls 2) compliance standards.

status · string · required

The current status of the device or job running on it.

---

Job History

jobs · array · required

An array of job entries representing each compute job executed on this device.

Show properties

jobs.compute_minutes_hired · integer · required

Total minutes a customer hired the device.

jobs.compute_minutes_served · integer · required

Total minutes the device was active during the job.

jobs.earned · integer · required

Total $IO earnings for this job.

jobs.slashed · integer · required

Earnings deducted (slashed) for downtime or failed performance.

jobs.total_hire_rate · integer · required

The full payment rate if the device achieved 100% uptime.

jobs.uptime_percent · integer · required

Percentage of uptime during the job.

jobs.start_time · datetime string · required

UTC timestamp when the job started.

jobs.end_time · datetime string · required

UTC timestamp when the job ended.

jobs.status · string · required

Job status (e.g., pending, running, completed).

jobs.for · string · required

Cluster ID associated with the job.

Example

{
  "jobs": [
    {
      "compute_minutes_hired": 300,
      "compute_minutes_served": 0,
      "earned": 0,
      "end_time": "2023-09-12 00:00:00",
      "for": "e242cd76-9e65-4de1-b3ec-1cb8b002b38d",
      "slashed": 0,
      "start_time": "2023-09-12 00:00:00",
      "status": "pending",
      "total_hire_rate": 10.8,
      "uptime_percent": 100
    }
  ]
}