Metrics

Query GPU utilization, memory, temperature, and power metrics

Get Workload Metrics

Retrieve GPU metrics for a specific workload.

GET /v1/workloads/{workload_id}/metrics

Path Parameters

Parameter	Type	Description
workload_id	string	The unique workload identifier

Query Parameters

Parameter	Type	Description
time_range	string	Time range: last_1h, last_6h, last_24h, job_lifetime (default: last_1h)
metrics	string	Comma-separated list of metrics to include

Available Metrics

Metric	Description
gpu_utilization	GPU compute utilization percentage
memory_utilization	GPU memory utilization percentage
temperature	GPU temperature in Celsius
power_usage	Power consumption in Watts
gpu_usage	Overall GPU usage metric

Example Request

curl -X GET "https://api.usechamber.io/v1/workloads/wl_abc123/metrics?time_range=last_6h&metrics=gpu_utilization,memory_utilization" \
  -H "Authorization: Bearer <token>" \
  -H "X-Organization-Id: <org-id>"

Example Response

{
  "data": {
    "workload_id": "wl_abc123",
    "time_range": "last_6h",
    "metrics": {
      "gpu_utilization": {
        "avg": 85.2,
        "min": 45.0,
        "max": 99.8,
        "current": 92.1
      },
      "memory_utilization": {
        "avg": 78.5,
        "min": 60.2,
        "max": 89.4,
        "current": 82.0
      }
    }
  },
  "request_id": "req_abc123"
}

---

Get Ranked Workload Metrics

Retrieve metrics across workloads, ranked by utilization or consumption.

GET /v1/workloads/metrics

Query Parameters

Parameter	Type	Description
sort_by	string	Sort metric: gpu_utilization, memory_utilization, power_usage, gpu_hours
order	string	Sort order: asc, desc (default: desc)
limit	integer	Number of results (1-100, default: 10)
initiative_id	string	Filter to specific team
status	string	Filter by workload status

Example Request

curl -X GET "https://api.usechamber.io/v1/workloads/metrics?sort_by=gpu_utilization&order=desc&limit=5" \
  -H "Authorization: Bearer <token>" \
  -H "X-Organization-Id: <org-id>"

Example Response

{
  "data": {
    "workloads": [
      {
        "workload_id": "wl_abc123",
        "name": "training-large-model",
        "gpu_utilization": {
          "avg": 95.2,
          "current": 97.1
        },
        "initiative_id": "init_ml_team"
      },
      {
        "workload_id": "wl_def456",
        "name": "inference-batch",
        "gpu_utilization": {
          "avg": 88.7,
          "current": 91.0
        },
        "initiative_id": "init_research"
      }
    ],
    "scope_average": {
      "gpu_utilization": 72.5
    }
  },
  "metadata": {
    "total_count": 45
  },
  "request_id": "req_abc123"
}