Debug
Raw JSON from every backend API route. Refresh to re-run.
GET /api/datadog/workloads
{
"status": 200,
"body": {
"workloads": [
{
"id": "11111111-1111-1111-1111-111111111111",
"name": "inference-prod-h100",
"gpuType": "H100",
"archetype": "steady_inference",
"tags": {
"workload": "inference-prod",
"env": "prod",
"team": "ml-serving",
"gpu_type": "H100"
}
},
{
"id": "22222222-2222-2222-2222-222222222222",
"name": "training-experiments-b200",
"gpuType": "B200",
"archetype": "bursty_training",
"tags": {
"workload": "training-experiments",
"env": "research",
"team": "ml-research",
"gpu_type": "B200"
}
},
{
"id": "33333333-3333-3333-3333-333333333333",
"name": "dev-notebooks-a100",
"gpuType": "A100",
"archetype": "interactive_dev",
"tags": {
"workload": "dev-notebooks",
"env": "dev",
"team": "platform",
"gpu_type": "A100"
}
}
]
}
}GET /api/ornn/curve?gpuType=H100&horizonDays=180
{
"status": 200,
"body": {
"gpuType": "H100",
"fetchedAt": "2026-10-23T00:18:27.211Z",
"horizonDays": 180,
"points": [
{
"t": "2026-10-23",
"price_usd_per_hour": 2.3743
},
{
"t": "2026-10-24",
"price_usd_per_hour": 2.3732
},
{
"t": "2026-10-25",
"price_usd_per_hour": 2.3716
},
{
"t": "2026-10-26",
"price_usd_per_hour": 2.3926
},
{
"t": "2026-10-27",
"price_usd_per_hour": 2.3952
},
{
"t": "2026-10-28",
"price_usd_per_hour": 2.3872
},
{
"t": "2026-10-29",
"price_usd_per_hour": 2.393
},
{
"t": "2026-10-30",
"price_usd_per_hour": 2.4083
},
{
"t": "2026-10-31",
"price_usd_per_hour": 2.4145
},
{
"t": "2026-11-01",
"price_usd_per_hour": 2.4209
},
{
"t": "2026-11-02",
"price_usd_per_hour": 2.4212
},
{
"t": "2026-11-03",
"price_usd_per_hour": 2.4099
},
{
"t": "2026-11-04",
"price_usd_per_hour": 2.4023
},
{
"t": "2026-11-05",
"price_usd_per_hour": 2.3909
},
{
"t": "2026-11-06",
"price_usd_per_hour": 2.4139
},
{
"t": "2026-11-07",
"price_usd_per_hour": 2.3881
},
{
"t": "2026-11-08",
"price_usd_per_hour": 2.4082
},
{
"t": "2026-11-09",
"price_usd_per_hour": 2.3941
},
{
"t": "2026-11-10",
"price_usd_per_hour": 2.3722
},
{
"t": "2026-11-11",
"price_usd_per_hour": 2.3722
},
{
"t": "2026-11-12",
"price_usd_per_hour": 2.3839
},
{
"t": "2026-11-13",
"price_usd_per_hour": 2.3989
},
{
"t": "2026-11-14",
"price_usd_per_hour": 2.4136
},
{
"t": "2026-11-15",
"price_usd_per_hour": 2.427
},
{
"t": "2026-11-16",
"price_usd_per_hour": 2.4469
},
{
"t": "2026-11-17",
"price_usd_per_hour": 2.4395
},
{
"t": "2026-11-18",
"price_usd_per_hour": 2.4438
},
{
"t": "2026-11-19",
"price_usd_per_hour": 2.4643
},
{
"t": "2026-11-20",
"price_usd_per_hour": 2.4722
},
{
"t": "2026-11-21",
"price_usd_per_hour": 2.4541
},
{
"t": "2026-11-22",
"price_usd_per_hour": 2.4515
},
{
"t": "2026-11-23",
"price_usd_per_hour": 2.4221
},
{
"t": "2026-11-24",
"price_usd_per_hour": 2.4074
},
{
"t": "2026-11-25",
"price_usd_per_hour": 2.4315
},
{
"t": "2026-11-26",
"price_usd_per_hour": 2.4391
},
{
"t": "2026-11-27",
"price_usd_per_hour": 2.4404
},
{
"t": "2026-11-28",
"price_usd_per_hour": 2.4391
},
{
"t": "2026-11-29",
"price_usd_per_hour": 2.4254
},
{
"t": "2026-11-30",
"price_usd_per_hour": 2.4266
},
{
"t": "2026-12-01",
"price_usd_per_hour": 2.408
},
{
"t": "2026-12-02",
"price_usd_per_hour": 2.3902
},
{
"t": "2026-12-03",
"price_usd_per_hour": 2.4048
},
{
"t": "2026-12-04",
"price_usd_per_hour": 2.4149
},
{
"t": "2026-12-05",
"price_usd_per_hour": 2.4149
},
{
"t": "2026-12-06",
"price_usd_per_hour": 2.4057
},
{
"t": "2026-12-07",
"price_usd_per_hour": 2.4211
},
{
"t": "2026-12-08",
"price_usd_per_hour": 2.4277
},
{
"t": "2026-12-09",
"price_usd_per_hour": 2.4086
},
{
"t": "2026-12-10",
"price_usd_per_hour": 2.3978
},
{
"t": "2026-12-11",
"price_usd_per_hour": 2.3992
},
{
"t": "2026-12-12",
"price_usd_per_hour": 2.3902
},
{
"t": "2026-12-13",
"price_usd_per_hour": 2.3901
},
{
"t": "2026-12-14",
"price_usd_per_hour": 2.3737
},
{
"t": "2026-12-15",
"price_usd_per_hour": 2.3832
},
{
"t": "2026-12-16",
"price_usd_per_hour": 2.352
},
{
"t": "2026-12-17",
"price_usd_per_hour": 2.3583
},
{
"t": "2026-12-18",
"price_usd_per_hour": 2.3641
},
{
"t": "2026-12-19",
"price_usd_per_hour": 2.3651
},
{
"t": "2026-12-20",
"price_usd_per_hour": 2.3596
},
{
"t": "2026-12-21",
"price_usd_per_hour": 2.3526
},
{
"t": "2026-12-22",
"price_usd_per_hour": 2.3498
},
{
"t": "2026-12-23",
"price_usd_per_hour": 2.344
},
{
"t": "2026-12-24",
"price_usd_per_hour": 2.3476
},
{
"t": "2026-12-25",
"price_usd_per_hour": 2.3325
},
{
"t": "2026-12-26",
"price_usd_per_hour": 2.3212
},
{
"t": "2026-12-27",
"price_usd_per_hour": 2.301
},
{
"t": "2026-12-28",
"price_usd_per_hour": 2.3036
},
{
"t": "2026-12-29",
"price_usd_per_hour": 2.2886
},
{
"t": "2026-12-30",
"price_usd_per_hour": 2.2719
},
{
"t": "2026-12-31",
"price_usd_per_hour": 2.2601
},
{
"t": "2027-01-01",
"price_usd_per_hour": 2.2551
},
{
"t": "2027-01-02",
"price_usd_per_hour": 2.249
},
{
"t": "2027-01-03",
"price_usd_per_hour": 2.255
},
{
"t": "2027-01-04",
"price_usd_per_hour": 2.2407
},
{
"t": "2027-01-05",
"price_usd_per_hour": 2.236
},
{
"t": "2027-01-06",
"price_usd_per_hour": 2.2406
},
{
"t": "2027-01-07",
"price_usd_per_hour": 2.2368
},
{
"t": "2027-01-08",
"price_usd_per_hour": 2.2505
},
{
"t": "2027-01-09",
"price_usd_per_hour": 2.2429
},
{
"t": "2027-01-10",
"price_usd_per_hour": 2.2437
},
{
"t": "2027-01-11",
"price_usd_per_hour": 2.2623
},
{
"t": "2027-01-12",
"price_usd_per_hour": 2.2644
},
{
"t": "2027-01-13",
"price_usd_per_hour": 2.2686
},
{
"t": "2027-01-14",
"price_usd_per_hour": 2.2743
},
{
"t": "2027-01-15",
"price_usd_per_hour": 2.265
},
{
"t": "2027-01-16",
"price_usd_per_hour": 2.2627
},
{
"t": "2027-01-17",
"price_usd_per_hour": 2.2548
},
{
"t": "2027-01-18",
"price_usd_per_hour": 2.2473
},
{
"t": "2027-01-19",
"price_usd_per_hour": 2.2481
},
{
"t": "2027-01-20",
"price_usd_per_hour": 2.2442
},
{
"t": "2027-01-21",
"price_usd_per_hour": 2.2534
},
{
"t": "2027-01-22",
"price_usd_per_hour": 2.2582
},
{
"t": "2027-01-23",
"price_usd_per_hour": 2.254
},
{
"t": "2027-01-24",
"price_usd_per_hour": 2.243
},
{
"t": "2027-01-25",
"price_usd_per_hour": 2.2208
},
{
"t": "2027-01-26",
"price_usd_per_hour": 2.2219
},
{
"t": "2027-01-27",
"price_usd_per_hour": 2.2146
},
{
"t": "2027-01-28",
"price_usd_per_hour": 2.2189
},
{
"t": "2027-01-29",
"price_usd_per_hour": 2.2364
},
{
"t": "2027-01-30",
"price_usd_per_hour": 2.2473
},
{
"t": "2027-01-31",
"price_usd_per_hour": 2.2432
},
{
"t": "2027-02-01",
"price_usd_per_hour": 2.2445
},
{
"t": "2027-02-02",
"price_usd_per_hour": 2.253
},
{
"t": "2027-02-03",
"price_usd_per_hour": 2.2634
},
{
"t": "2027-02-04",
"price_usd_per_hour": 2.2803
},
{
"t": "2027-02-05",
"price_usd_per_hour": 2.2727
},
{
"t": "2027-02-06",
"price_usd_per_hour": 2.2656
},
{
"t": "2027-02-07",
"price_usd_per_hour": 2.2529
},
{
"t": "2027-02-08",
"price_usd_per_hour": 2.2424
},
{
"t": "2027-02-09",
"price_usd_per_hour": 2.2344
},
{
"t": "2027-02-10",
"price_usd_per_hour": 2.2309
},
{
"t": "2027-02-11",
"price_usd_per_hour": 2.2254
},
{
"t": "2027-02-12",
"price_usd_per_hour": 2.2138
},
{
"t": "2027-02-13",
"price_usd_per_hour": 2.2085
},
{
"t": "2027-02-14",
"price_usd_per_hour": 2.1995
},
{
"t": "2027-02-15",
"price_usd_per_hour": 2.1914
},
{
"t": "2027-02-16",
"price_usd_per_hour": 2.1874
},
{
"t": "2027-02-17",
"price_usd_per_hour": 2.1946
},
{
"t": "2027-02-18",
"price_usd_per_hour": 2.2083
},
{
"t": "2027-02-19",
"price_usd_per_hour": 2.2037
},
{
"t": "2027-02-20",
"price_usd_per_hour": 2.1778
},
{
"t": "2027-02-21",
"price_usd_per_hour": 2.1397
},
{
"t": "2027-02-22",
"price_usd_per_hour": 2.1235
},
{
"t": "2027-02-23",
"price_usd_per_hour": 2.1166
},
{
"t": "2027-02-24",
"price_usd_per_hour": 2.1062
},
{
"t": "2027-02-25",
"price_usd_per_hour": 2.0984
},
{
"t": "2027-02-26",
"price_usd_per_hour": 2.0893
},
{
"t": "2027-02-27",
"price_usd_per_hour": 2.0801
},
{
"t": "2027-02-28",
"price_usd_per_hour": 2.0921
},
{
"t": "2027-03-01",
"price_usd_per_hour": 2.0586
},
{
"t": "2027-03-02",
"price_usd_per_hour": 2.0534
},
{
"t": "2027-03-03",
"price_usd_per_hour": 2.0582
},
{
"t": "2027-03-04",
"price_usd_per_hour": 2.0463
},
{
"t": "2027-03-05",
"price_usd_per_hour": 2.0596
},
{
"t": "2027-03-06",
"price_usd_per_hour": 2.0599
},
{
"t": "2027-03-07",
"price_usd_per_hour": 2.0555
},
{
"t": "2027-03-08",
"price_usd_per_hour": 2.0524
},
{
"t": "2027-03-09",
"price_usd_per_hour": 2.054
},
{
"t": "2027-03-10",
"price_usd_per_hour": 2.0456
},
{
"t": "2027-03-11",
"price_usd_per_hour": 2.0462
},
{
"t": "2027-03-12",
"price_usd_per_hour": 2.0403
},
{
"t": "2027-03-13",
"price_usd_per_hour": 2.0682
},
{
"t": "2027-03-14",
"price_usd_per_hour": 2.0742
},
{
"t": "2027-03-15",
"price_usd_per_hour": 2.0891
},
{
"t": "2027-03-16",
"price_usd_per_hour": 2.0987
},
{
"t": "2027-03-17",
"price_usd_per_hour": 2.1035
},
{
"t": "2027-03-18",
"price_usd_per_hour": 2.099
},
{
"t": "2027-03-19",
"price_usd_per_hour": 2.0782
},
{
"t": "2027-03-20",
"price_usd_per_hour": 2.0576
},
{
"t": "2027-03-21",
"price_usd_per_hour": 2.036
},
{
"t": "2027-03-22",
"price_usd_per_hour": 2.0269
},
{
"t": "2027-03-23",
"price_usd_per_hour": 2.0181
},
{
"t": "2027-03-24",
"price_usd_per_hour": 1.9932
},
{
"t": "2027-03-25",
"price_usd_per_hour": 1.9971
},
{
"t": "2027-03-26",
"price_usd_per_hour": 2.0028
},
{
"t": "2027-03-27",
"price_usd_per_hour": 2.0124
},
{
"t": "2027-03-28",
"price_usd_per_hour": 2.0228
},
{
"t": "2027-03-29",
"price_usd_per_hour": 2.0185
},
{
"t": "2027-03-30",
"price_usd_per_hour": 2.0156
},
{
"t": "2027-03-31",
"price_usd_per_hour": 1.9949
},
{
"t": "2027-04-01",
"price_usd_per_hour": 2.0069
},
{
"t": "2027-04-02",
"price_usd_per_hour": 2.0221
},
{
"t": "2027-04-03",
"price_usd_per_hour": 2.0147
},
{
"t": "2027-04-04",
"price_usd_per_hour": 2.0202
},
{
"t": "2027-04-05",
"price_usd_per_hour": 2.0289
},
{
"t": "2027-04-06",
"price_usd_per_hour": 2.0403
},
{
"t": "2027-04-07",
"price_usd_per_hour": 2.0141
},
{
"t": "2027-04-08",
"price_usd_per_hour": 2.0104
},
{
"t": "2027-04-09",
"price_usd_per_hour": 2.0227
},
{
"t": "2027-04-10",
"price_usd_per_hour": 2.0086
},
{
"t": "2027-04-11",
"price_usd_per_hour": 2.0094
},
{
"t": "2027-04-12",
"price_usd_per_hour": 2.009
},
{
"t": "2027-04-13",
"price_usd_per_hour": 2.0056
},
{
"t": "2027-04-14",
"price_usd_per_hour": 2.0051
},
{
"t": "2027-04-15",
"price_usd_per_hour": 1.9923
},
{
"t": "2027-04-16",
"price_usd_per_hour": 1.9894
},
{
"t": "2027-04-17",
"price_usd_per_hour": 1.9673
},
{
"t": "2027-04-18",
"price_usd_per_hour": 1.9751
},
{
"t": "2027-04-19",
"price_usd_per_hour": 1.9686
},
{
"t": "2027-04-20",
"price_usd_per_hour": 1.9746
}
],
"source": "ornn_fixture",
"persisted": false
}
}GET /api/datadog/usage (last 120d, first workload)
{
"status": 200,
"body": {
"workloadId": "11111111-1111-1111-1111-111111111111",
"from": "2026-06-25",
"to": "2026-10-23",
"daily": [
{
"day": "2026-06-25",
"gpuHours": 331.27,
"onDemandHours": 111.27,
"reservedHours": 220,
"costUsd": 3962.51,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-06-26",
"gpuHours": 331.2,
"onDemandHours": 111.2,
"reservedHours": 220,
"costUsd": 3936.97,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-06-27",
"gpuHours": 268.52,
"onDemandHours": 48.52,
"reservedHours": 220,
"costUsd": 3127.9,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-06-28",
"gpuHours": 270.42,
"onDemandHours": 50.42,
"reservedHours": 220,
"costUsd": 2957.06,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-06-29",
"gpuHours": 269.89,
"onDemandHours": 49.89,
"reservedHours": 220,
"costUsd": 3190.37,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-06-30",
"gpuHours": 314.39,
"onDemandHours": 94.39,
"reservedHours": 220,
"costUsd": 3675.1,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-01",
"gpuHours": 271.36,
"onDemandHours": 51.36,
"reservedHours": 220,
"costUsd": 3057.92,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-02",
"gpuHours": 283.06,
"onDemandHours": 63.06,
"reservedHours": 220,
"costUsd": 3154.1,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-03",
"gpuHours": 309.95,
"onDemandHours": 89.95,
"reservedHours": 220,
"costUsd": 3490.36,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-04",
"gpuHours": 277.75,
"onDemandHours": 57.75,
"reservedHours": 220,
"costUsd": 3104.13,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-05",
"gpuHours": 260.29,
"onDemandHours": 40.29,
"reservedHours": 220,
"costUsd": 2814.16,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-06",
"gpuHours": 298.01,
"onDemandHours": 78.01,
"reservedHours": 220,
"costUsd": 3343.11,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-07",
"gpuHours": 273.42,
"onDemandHours": 53.42,
"reservedHours": 220,
"costUsd": 3168.5,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-08",
"gpuHours": 325.47,
"onDemandHours": 105.47,
"reservedHours": 220,
"costUsd": 3641.89,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-09",
"gpuHours": 279.8,
"onDemandHours": 59.8,
"reservedHours": 220,
"costUsd": 3214.01,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-10",
"gpuHours": 296.77,
"onDemandHours": 76.77,
"reservedHours": 220,
"costUsd": 3413.77,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-11",
"gpuHours": 272.8,
"onDemandHours": 52.8,
"reservedHours": 220,
"costUsd": 3146.28,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-12",
"gpuHours": 289.99,
"onDemandHours": 69.99,
"reservedHours": 220,
"costUsd": 3158.71,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-13",
"gpuHours": 294,
"onDemandHours": 74,
"reservedHours": 220,
"costUsd": 3205.32,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-14",
"gpuHours": 283.39,
"onDemandHours": 63.39,
"reservedHours": 220,
"costUsd": 3126.64,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-15",
"gpuHours": 267.37,
"onDemandHours": 47.37,
"reservedHours": 220,
"costUsd": 2884.18,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-16",
"gpuHours": 280.64,
"onDemandHours": 60.64,
"reservedHours": 220,
"costUsd": 3133.57,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-17",
"gpuHours": 300.76,
"onDemandHours": 80.76,
"reservedHours": 220,
"costUsd": 3467.5,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-18",
"gpuHours": 241.05,
"onDemandHours": 21.05,
"reservedHours": 220,
"costUsd": 2531.3,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-19",
"gpuHours": 268.97,
"onDemandHours": 48.97,
"reservedHours": 220,
"costUsd": 2945.18,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-20",
"gpuHours": 267.93,
"onDemandHours": 47.93,
"reservedHours": 220,
"costUsd": 2809.77,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-21",
"gpuHours": 257.9,
"onDemandHours": 37.9,
"reservedHours": 220,
"costUsd": 2748.26,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-22",
"gpuHours": 222.34,
"onDemandHours": 2.34,
"reservedHours": 220,
"costUsd": 2114.44,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-23",
"gpuHours": 274.41,
"onDemandHours": 54.41,
"reservedHours": 220,
"costUsd": 3051.8,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-24",
"gpuHours": 269.64,
"onDemandHours": 49.64,
"reservedHours": 220,
"costUsd": 3057.15,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-25",
"gpuHours": 244.45,
"onDemandHours": 24.45,
"reservedHours": 220,
"costUsd": 2602.74,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-26",
"gpuHours": 247.6,
"onDemandHours": 27.6,
"reservedHours": 220,
"costUsd": 2618.81,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-27",
"gpuHours": 315.84,
"onDemandHours": 95.84,
"reservedHours": 220,
"costUsd": 3763.31,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-28",
"gpuHours": 300.19,
"onDemandHours": 80.19,
"reservedHours": 220,
"costUsd": 3466.02,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-29",
"gpuHours": 285.84,
"onDemandHours": 65.84,
"reservedHours": 220,
"costUsd": 3154.42,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-30",
"gpuHours": 243.49,
"onDemandHours": 23.49,
"reservedHours": 220,
"costUsd": 2634.67,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-07-31",
"gpuHours": 233.46,
"onDemandHours": 13.46,
"reservedHours": 220,
"costUsd": 2428.52,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-01",
"gpuHours": 222.83,
"onDemandHours": 2.83,
"reservedHours": 220,
"costUsd": 2294.41,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-02",
"gpuHours": 232.82,
"onDemandHours": 12.82,
"reservedHours": 220,
"costUsd": 2311.53,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-03",
"gpuHours": 239.4,
"onDemandHours": 19.4,
"reservedHours": 220,
"costUsd": 2517.38,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-04",
"gpuHours": 258.34,
"onDemandHours": 38.34,
"reservedHours": 220,
"costUsd": 2648.13,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-05",
"gpuHours": 264.81,
"onDemandHours": 44.81,
"reservedHours": 220,
"costUsd": 2857.04,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-06",
"gpuHours": 311.86,
"onDemandHours": 91.86,
"reservedHours": 220,
"costUsd": 3755.45,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-07",
"gpuHours": 256.82,
"onDemandHours": 36.82,
"reservedHours": 220,
"costUsd": 2776.01,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-08",
"gpuHours": 252.86,
"onDemandHours": 32.86,
"reservedHours": 220,
"costUsd": 2681.1,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-09",
"gpuHours": 197.21,
"onDemandHours": 0,
"reservedHours": 197.21,
"costUsd": 1908.2,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-10",
"gpuHours": 232,
"onDemandHours": 12,
"reservedHours": 220,
"costUsd": 2507.68,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-11",
"gpuHours": 248.38,
"onDemandHours": 28.38,
"reservedHours": 220,
"costUsd": 2635.68,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-12",
"gpuHours": 300.62,
"onDemandHours": 80.62,
"reservedHours": 220,
"costUsd": 3445.2,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-13",
"gpuHours": 254.24,
"onDemandHours": 34.24,
"reservedHours": 220,
"costUsd": 2762.81,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-14",
"gpuHours": 265.37,
"onDemandHours": 45.37,
"reservedHours": 220,
"costUsd": 2960.86,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-15",
"gpuHours": 232.71,
"onDemandHours": 12.71,
"reservedHours": 220,
"costUsd": 2370.17,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-16",
"gpuHours": 230.39,
"onDemandHours": 10.39,
"reservedHours": 220,
"costUsd": 2352.87,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-17",
"gpuHours": 269.15,
"onDemandHours": 49.15,
"reservedHours": 220,
"costUsd": 3002.27,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-18",
"gpuHours": 218.97,
"onDemandHours": 0,
"reservedHours": 218.97,
"costUsd": 2240.18,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-19",
"gpuHours": 264.82,
"onDemandHours": 44.82,
"reservedHours": 220,
"costUsd": 2957.05,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-20",
"gpuHours": 250.62,
"onDemandHours": 30.62,
"reservedHours": 220,
"costUsd": 2680.54,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-21",
"gpuHours": 261.47,
"onDemandHours": 41.47,
"reservedHours": 220,
"costUsd": 2850.94,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-22",
"gpuHours": 205.11,
"onDemandHours": 0,
"reservedHours": 205.11,
"costUsd": 2082.92,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-23",
"gpuHours": 238.89,
"onDemandHours": 18.89,
"reservedHours": 220,
"costUsd": 2547.96,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-24",
"gpuHours": 280.02,
"onDemandHours": 60.02,
"reservedHours": 220,
"costUsd": 3342.37,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-25",
"gpuHours": 282.36,
"onDemandHours": 62.36,
"reservedHours": 220,
"costUsd": 3191.1,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-26",
"gpuHours": 273.67,
"onDemandHours": 53.67,
"reservedHours": 220,
"costUsd": 3117.97,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-27",
"gpuHours": 281.1,
"onDemandHours": 61.1,
"reservedHours": 220,
"costUsd": 3115.3,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-28",
"gpuHours": 288.02,
"onDemandHours": 68.02,
"reservedHours": 220,
"costUsd": 3507.03,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-29",
"gpuHours": 214.38,
"onDemandHours": 0,
"reservedHours": 214.38,
"costUsd": 2216.12,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-30",
"gpuHours": 254.18,
"onDemandHours": 34.18,
"reservedHours": 220,
"costUsd": 2787.39,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-08-31",
"gpuHours": 285.42,
"onDemandHours": 65.42,
"reservedHours": 220,
"costUsd": 3214.84,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-01",
"gpuHours": 278.94,
"onDemandHours": 58.94,
"reservedHours": 220,
"costUsd": 3168.71,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-02",
"gpuHours": 268,
"onDemandHours": 48,
"reservedHours": 220,
"costUsd": 3114.7,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-03",
"gpuHours": 258.56,
"onDemandHours": 38.56,
"reservedHours": 220,
"costUsd": 2941.45,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-04",
"gpuHours": 314.46,
"onDemandHours": 94.46,
"reservedHours": 220,
"costUsd": 3729.93,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-05",
"gpuHours": 243.09,
"onDemandHours": 23.09,
"reservedHours": 220,
"costUsd": 2544.57,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-06",
"gpuHours": 269.01,
"onDemandHours": 49.01,
"reservedHours": 220,
"costUsd": 2997.28,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-07",
"gpuHours": 300.52,
"onDemandHours": 80.52,
"reservedHours": 220,
"costUsd": 3668.84,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-08",
"gpuHours": 299.7,
"onDemandHours": 79.7,
"reservedHours": 220,
"costUsd": 3610.82,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-09",
"gpuHours": 262.77,
"onDemandHours": 42.77,
"reservedHours": 220,
"costUsd": 2936.39,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-10",
"gpuHours": 238.86,
"onDemandHours": 18.86,
"reservedHours": 220,
"costUsd": 2580.66,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-11",
"gpuHours": 270.91,
"onDemandHours": 50.91,
"reservedHours": 220,
"costUsd": 3056.38,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-12",
"gpuHours": 236.58,
"onDemandHours": 16.58,
"reservedHours": 220,
"costUsd": 2554.74,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-13",
"gpuHours": 251.35,
"onDemandHours": 31.35,
"reservedHours": 220,
"costUsd": 2767.35,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-14",
"gpuHours": 245.16,
"onDemandHours": 25.16,
"reservedHours": 220,
"costUsd": 2741.02,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-15",
"gpuHours": 316.79,
"onDemandHours": 96.79,
"reservedHours": 220,
"costUsd": 3824.64,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-16",
"gpuHours": 284.44,
"onDemandHours": 64.44,
"reservedHours": 220,
"costUsd": 3333.11,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-17",
"gpuHours": 289.26,
"onDemandHours": 69.26,
"reservedHours": 220,
"costUsd": 3568.44,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-18",
"gpuHours": 267.5,
"onDemandHours": 47.5,
"reservedHours": 220,
"costUsd": 3278.7,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-19",
"gpuHours": 217.96,
"onDemandHours": 0,
"reservedHours": 217.96,
"costUsd": 2252.61,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-20",
"gpuHours": 245.11,
"onDemandHours": 25.11,
"reservedHours": 220,
"costUsd": 2721.57,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-21",
"gpuHours": 283.17,
"onDemandHours": 63.17,
"reservedHours": 220,
"costUsd": 3391.52,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-22",
"gpuHours": 277.55,
"onDemandHours": 57.55,
"reservedHours": 220,
"costUsd": 3267.05,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-23",
"gpuHours": 269.08,
"onDemandHours": 49.08,
"reservedHours": 220,
"costUsd": 3223.32,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-24",
"gpuHours": 303.23,
"onDemandHours": 83.23,
"reservedHours": 220,
"costUsd": 3808.99,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-25",
"gpuHours": 263.7,
"onDemandHours": 43.7,
"reservedHours": 220,
"costUsd": 3062.73,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-26",
"gpuHours": 225.29,
"onDemandHours": 5.29,
"reservedHours": 220,
"costUsd": 2372.14,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-27",
"gpuHours": 255.61,
"onDemandHours": 35.61,
"reservedHours": 220,
"costUsd": 3122.31,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-28",
"gpuHours": 274.35,
"onDemandHours": 54.35,
"reservedHours": 220,
"costUsd": 3490.33,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-29",
"gpuHours": 234.95,
"onDemandHours": 14.95,
"reservedHours": 220,
"costUsd": 2729.31,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-09-30",
"gpuHours": 291.12,
"onDemandHours": 71.12,
"reservedHours": 220,
"costUsd": 3612.03,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-01",
"gpuHours": 236.04,
"onDemandHours": 16.04,
"reservedHours": 220,
"costUsd": 2828.47,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-02",
"gpuHours": 262.96,
"onDemandHours": 42.96,
"reservedHours": 220,
"costUsd": 3052,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-03",
"gpuHours": 260.12,
"onDemandHours": 40.12,
"reservedHours": 220,
"costUsd": 2981.8,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-04",
"gpuHours": 265.62,
"onDemandHours": 45.62,
"reservedHours": 220,
"costUsd": 3111.69,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-05",
"gpuHours": 294.57,
"onDemandHours": 74.57,
"reservedHours": 220,
"costUsd": 3480.66,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-06",
"gpuHours": 289.36,
"onDemandHours": 69.36,
"reservedHours": 220,
"costUsd": 3350.48,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-07",
"gpuHours": 245.19,
"onDemandHours": 25.19,
"reservedHours": 220,
"costUsd": 2769.41,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-08",
"gpuHours": 279.88,
"onDemandHours": 59.88,
"reservedHours": 220,
"costUsd": 3428.68,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-09",
"gpuHours": 304.23,
"onDemandHours": 84.23,
"reservedHours": 220,
"costUsd": 3743.69,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-10",
"gpuHours": 235.78,
"onDemandHours": 15.78,
"reservedHours": 220,
"costUsd": 2493.4,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-11",
"gpuHours": 252.5,
"onDemandHours": 32.5,
"reservedHours": 220,
"costUsd": 2955.11,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-12",
"gpuHours": 268.97,
"onDemandHours": 48.97,
"reservedHours": 220,
"costUsd": 3245.42,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-13",
"gpuHours": 228.22,
"onDemandHours": 8.22,
"reservedHours": 220,
"costUsd": 2582.81,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-14",
"gpuHours": 257.07,
"onDemandHours": 37.07,
"reservedHours": 220,
"costUsd": 2977.1,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-15",
"gpuHours": 267.36,
"onDemandHours": 47.36,
"reservedHours": 220,
"costUsd": 3317.26,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-16",
"gpuHours": 262.37,
"onDemandHours": 42.37,
"reservedHours": 220,
"costUsd": 2990.87,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-17",
"gpuHours": 260.85,
"onDemandHours": 40.85,
"reservedHours": 220,
"costUsd": 2892.93,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-18",
"gpuHours": 231.42,
"onDemandHours": 11.42,
"reservedHours": 220,
"costUsd": 2627.76,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-19",
"gpuHours": 261.47,
"onDemandHours": 41.47,
"reservedHours": 220,
"costUsd": 3187.54,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-20",
"gpuHours": 315.76,
"onDemandHours": 95.76,
"reservedHours": 220,
"costUsd": 3827.65,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-21",
"gpuHours": 286.89,
"onDemandHours": 66.89,
"reservedHours": 220,
"costUsd": 3571.04,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-22",
"gpuHours": 255.24,
"onDemandHours": 35.24,
"reservedHours": 220,
"costUsd": 2916.41,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
},
{
"day": "2026-10-23",
"gpuHours": 270.51,
"onDemandHours": 50.51,
"reservedHours": 220,
"costUsd": 3301.46,
"provenance": {
"utilization_query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"cost_query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
}
}
],
"series": {
"workloadId": "11111111-1111-1111-1111-111111111111",
"from": "2026-06-25",
"to": "2026-10-23",
"gpuUtilization": {
"query": "avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"unit": "percent",
"pointlist": [
[
1782345600000,
0.8282
],
[
1782432000000,
0.828
],
[
1782518400000,
0.6713
],
[
1782604800000,
0.676
],
[
1782691200000,
0.6747
],
[
1782777600000,
0.786
],
[
1782864000000,
0.6784
],
[
1782950400000,
0.7076
],
[
1783036800000,
0.7749
],
[
1783123200000,
0.6944
],
[
1783209600000,
0.6507
],
[
1783296000000,
0.745
],
[
1783382400000,
0.6836
],
[
1783468800000,
0.8137
],
[
1783555200000,
0.6995
],
[
1783641600000,
0.7419
],
[
1783728000000,
0.682
],
[
1783814400000,
0.725
],
[
1783900800000,
0.735
],
[
1783987200000,
0.7085
],
[
1784073600000,
0.6684
],
[
1784160000000,
0.7016
],
[
1784246400000,
0.7519
],
[
1784332800000,
0.6026
],
[
1784419200000,
0.6724
],
[
1784505600000,
0.6698
],
[
1784592000000,
0.6447
],
[
1784678400000,
0.5559
],
[
1784764800000,
0.686
],
[
1784851200000,
0.6741
],
[
1784937600000,
0.6111
],
[
1785024000000,
0.619
],
[
1785110400000,
0.7896
],
[
1785196800000,
0.7505
],
[
1785283200000,
0.7146
],
[
1785369600000,
0.6087
],
[
1785456000000,
0.5836
],
[
1785542400000,
0.5571
],
[
1785628800000,
0.582
],
[
1785715200000,
0.5985
],
[
1785801600000,
0.6458
],
[
1785888000000,
0.662
],
[
1785974400000,
0.7797
],
[
1786060800000,
0.642
],
[
1786147200000,
0.6322
],
[
1786233600000,
0.493
],
[
1786320000000,
0.58
],
[
1786406400000,
0.6209
],
[
1786492800000,
0.7515
],
[
1786579200000,
0.6356
],
[
1786665600000,
0.6634
],
[
1786752000000,
0.5818
],
[
1786838400000,
0.576
],
[
1786924800000,
0.6729
],
[
1787011200000,
0.5474
],
[
1787097600000,
0.6621
],
[
1787184000000,
0.6266
],
[
1787270400000,
0.6537
],
[
1787356800000,
0.5128
],
[
1787443200000,
0.5972
],
[
1787529600000,
0.7001
],
[
1787616000000,
0.7059
],
[
1787702400000,
0.6842
],
[
1787788800000,
0.7028
],
[
1787875200000,
0.7201
],
[
1787961600000,
0.5359
],
[
1788048000000,
0.6355
],
[
1788134400000,
0.7135
],
[
1788220800000,
0.6974
],
[
1788307200000,
0.67
],
[
1788393600000,
0.6464
],
[
1788480000000,
0.7862
],
[
1788566400000,
0.6077
],
[
1788652800000,
0.6725
],
[
1788739200000,
0.7513
],
[
1788825600000,
0.7492
],
[
1788912000000,
0.6569
],
[
1788998400000,
0.5972
],
[
1789084800000,
0.6773
],
[
1789171200000,
0.5914
],
[
1789257600000,
0.6284
],
[
1789344000000,
0.6129
],
[
1789430400000,
0.792
],
[
1789516800000,
0.7111
],
[
1789603200000,
0.7232
],
[
1789689600000,
0.6688
],
[
1789776000000,
0.5449
],
[
1789862400000,
0.6128
],
[
1789948800000,
0.7079
],
[
1790035200000,
0.6939
],
[
1790121600000,
0.6727
],
[
1790208000000,
0.7581
],
[
1790294400000,
0.6593
],
[
1790380800000,
0.5632
],
[
1790467200000,
0.639
],
[
1790553600000,
0.6859
],
[
1790640000000,
0.5874
],
[
1790726400000,
0.7278
],
[
1790812800000,
0.5901
],
[
1790899200000,
0.6574
],
[
1790985600000,
0.6503
],
[
1791072000000,
0.6641
],
[
1791158400000,
0.7364
],
[
1791244800000,
0.7234
],
[
1791331200000,
0.613
],
[
1791417600000,
0.6997
],
[
1791504000000,
0.7606
],
[
1791590400000,
0.5895
],
[
1791676800000,
0.6313
],
[
1791763200000,
0.6724
],
[
1791849600000,
0.5706
],
[
1791936000000,
0.6427
],
[
1792022400000,
0.6684
],
[
1792108800000,
0.6559
],
[
1792195200000,
0.6521
],
[
1792281600000,
0.5786
],
[
1792368000000,
0.6537
],
[
1792454400000,
0.7894
],
[
1792540800000,
0.7172
],
[
1792627200000,
0.6381
],
[
1792713600000,
0.6763
]
]
},
"gpuHours": {
"query": "sum:system.gpu.hours{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)",
"unit": "gpu_hour",
"pointlist": [
[
1782345600000,
331.27
],
[
1782432000000,
331.2
],
[
1782518400000,
268.52
],
[
1782604800000,
270.42
],
[
1782691200000,
269.89
],
[
1782777600000,
314.39
],
[
1782864000000,
271.36
],
[
1782950400000,
283.06
],
[
1783036800000,
309.95
],
[
1783123200000,
277.75
],
[
1783209600000,
260.29
],
[
1783296000000,
298.01
],
[
1783382400000,
273.42
],
[
1783468800000,
325.47
],
[
1783555200000,
279.8
],
[
1783641600000,
296.77
],
[
1783728000000,
272.8
],
[
1783814400000,
289.99
],
[
1783900800000,
294
],
[
1783987200000,
283.39
],
[
1784073600000,
267.37
],
[
1784160000000,
280.64
],
[
1784246400000,
300.76
],
[
1784332800000,
241.05
],
[
1784419200000,
268.97
],
[
1784505600000,
267.93
],
[
1784592000000,
257.9
],
[
1784678400000,
222.34
],
[
1784764800000,
274.41
],
[
1784851200000,
269.64
],
[
1784937600000,
244.45
],
[
1785024000000,
247.6
],
[
1785110400000,
315.84
],
[
1785196800000,
300.19
],
[
1785283200000,
285.84
],
[
1785369600000,
243.49
],
[
1785456000000,
233.46
],
[
1785542400000,
222.83
],
[
1785628800000,
232.82
],
[
1785715200000,
239.4
],
[
1785801600000,
258.34
],
[
1785888000000,
264.81
],
[
1785974400000,
311.86
],
[
1786060800000,
256.82
],
[
1786147200000,
252.86
],
[
1786233600000,
197.21
],
[
1786320000000,
232
],
[
1786406400000,
248.38
],
[
1786492800000,
300.62
],
[
1786579200000,
254.24
],
[
1786665600000,
265.37
],
[
1786752000000,
232.71
],
[
1786838400000,
230.39
],
[
1786924800000,
269.15
],
[
1787011200000,
218.97
],
[
1787097600000,
264.82
],
[
1787184000000,
250.62
],
[
1787270400000,
261.47
],
[
1787356800000,
205.11
],
[
1787443200000,
238.89
],
[
1787529600000,
280.02
],
[
1787616000000,
282.36
],
[
1787702400000,
273.67
],
[
1787788800000,
281.1
],
[
1787875200000,
288.02
],
[
1787961600000,
214.38
],
[
1788048000000,
254.18
],
[
1788134400000,
285.42
],
[
1788220800000,
278.94
],
[
1788307200000,
268
],
[
1788393600000,
258.56
],
[
1788480000000,
314.46
],
[
1788566400000,
243.09
],
[
1788652800000,
269.01
],
[
1788739200000,
300.52
],
[
1788825600000,
299.7
],
[
1788912000000,
262.77
],
[
1788998400000,
238.86
],
[
1789084800000,
270.91
],
[
1789171200000,
236.58
],
[
1789257600000,
251.35
],
[
1789344000000,
245.16
],
[
1789430400000,
316.79
],
[
1789516800000,
284.44
],
[
1789603200000,
289.26
],
[
1789689600000,
267.5
],
[
1789776000000,
217.96
],
[
1789862400000,
245.11
],
[
1789948800000,
283.17
],
[
1790035200000,
277.55
],
[
1790121600000,
269.08
],
[
1790208000000,
303.23
],
[
1790294400000,
263.7
],
[
1790380800000,
225.29
],
[
1790467200000,
255.61
],
[
1790553600000,
274.35
],
[
1790640000000,
234.95
],
[
1790726400000,
291.12
],
[
1790812800000,
236.04
],
[
1790899200000,
262.96
],
[
1790985600000,
260.12
],
[
1791072000000,
265.62
],
[
1791158400000,
294.57
],
[
1791244800000,
289.36
],
[
1791331200000,
245.19
],
[
1791417600000,
279.88
],
[
1791504000000,
304.23
],
[
1791590400000,
235.78
],
[
1791676800000,
252.5
],
[
1791763200000,
268.97
],
[
1791849600000,
228.22
],
[
1791936000000,
257.07
],
[
1792022400000,
267.36
],
[
1792108800000,
262.37
],
[
1792195200000,
260.85
],
[
1792281600000,
231.42
],
[
1792368000000,
261.47
],
[
1792454400000,
315.76
],
[
1792540800000,
286.89
],
[
1792627200000,
255.24
],
[
1792713600000,
270.51
]
]
},
"onDemandHours": {
"query": "sum:aws.compute.gpu_hours{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100,pricing:ondemand}.rollup(sum, 86400)",
"unit": "gpu_hour",
"pointlist": [
[
1782345600000,
111.27
],
[
1782432000000,
111.2
],
[
1782518400000,
48.52
],
[
1782604800000,
50.42
],
[
1782691200000,
49.89
],
[
1782777600000,
94.39
],
[
1782864000000,
51.36
],
[
1782950400000,
63.06
],
[
1783036800000,
89.95
],
[
1783123200000,
57.75
],
[
1783209600000,
40.29
],
[
1783296000000,
78.01
],
[
1783382400000,
53.42
],
[
1783468800000,
105.47
],
[
1783555200000,
59.8
],
[
1783641600000,
76.77
],
[
1783728000000,
52.8
],
[
1783814400000,
69.99
],
[
1783900800000,
74
],
[
1783987200000,
63.39
],
[
1784073600000,
47.37
],
[
1784160000000,
60.64
],
[
1784246400000,
80.76
],
[
1784332800000,
21.05
],
[
1784419200000,
48.97
],
[
1784505600000,
47.93
],
[
1784592000000,
37.9
],
[
1784678400000,
2.34
],
[
1784764800000,
54.41
],
[
1784851200000,
49.64
],
[
1784937600000,
24.45
],
[
1785024000000,
27.6
],
[
1785110400000,
95.84
],
[
1785196800000,
80.19
],
[
1785283200000,
65.84
],
[
1785369600000,
23.49
],
[
1785456000000,
13.46
],
[
1785542400000,
2.83
],
[
1785628800000,
12.82
],
[
1785715200000,
19.4
],
[
1785801600000,
38.34
],
[
1785888000000,
44.81
],
[
1785974400000,
91.86
],
[
1786060800000,
36.82
],
[
1786147200000,
32.86
],
[
1786233600000,
0
],
[
1786320000000,
12
],
[
1786406400000,
28.38
],
[
1786492800000,
80.62
],
[
1786579200000,
34.24
],
[
1786665600000,
45.37
],
[
1786752000000,
12.71
],
[
1786838400000,
10.39
],
[
1786924800000,
49.15
],
[
1787011200000,
0
],
[
1787097600000,
44.82
],
[
1787184000000,
30.62
],
[
1787270400000,
41.47
],
[
1787356800000,
0
],
[
1787443200000,
18.89
],
[
1787529600000,
60.02
],
[
1787616000000,
62.36
],
[
1787702400000,
53.67
],
[
1787788800000,
61.1
],
[
1787875200000,
68.02
],
[
1787961600000,
0
],
[
1788048000000,
34.18
],
[
1788134400000,
65.42
],
[
1788220800000,
58.94
],
[
1788307200000,
48
],
[
1788393600000,
38.56
],
[
1788480000000,
94.46
],
[
1788566400000,
23.09
],
[
1788652800000,
49.01
],
[
1788739200000,
80.52
],
[
1788825600000,
79.7
],
[
1788912000000,
42.77
],
[
1788998400000,
18.86
],
[
1789084800000,
50.91
],
[
1789171200000,
16.58
],
[
1789257600000,
31.35
],
[
1789344000000,
25.16
],
[
1789430400000,
96.79
],
[
1789516800000,
64.44
],
[
1789603200000,
69.26
],
[
1789689600000,
47.5
],
[
1789776000000,
0
],
[
1789862400000,
25.11
],
[
1789948800000,
63.17
],
[
1790035200000,
57.55
],
[
1790121600000,
49.08
],
[
1790208000000,
83.23
],
[
1790294400000,
43.7
],
[
1790380800000,
5.29
],
[
1790467200000,
35.61
],
[
1790553600000,
54.35
],
[
1790640000000,
14.95
],
[
1790726400000,
71.12
],
[
1790812800000,
16.04
],
[
1790899200000,
42.96
],
[
1790985600000,
40.12
],
[
1791072000000,
45.62
],
[
1791158400000,
74.57
],
[
1791244800000,
69.36
],
[
1791331200000,
25.19
],
[
1791417600000,
59.88
],
[
1791504000000,
84.23
],
[
1791590400000,
15.78
],
[
1791676800000,
32.5
],
[
1791763200000,
48.97
],
[
1791849600000,
8.22
],
[
1791936000000,
37.07
],
[
1792022400000,
47.36
],
[
1792108800000,
42.37
],
[
1792195200000,
40.85
],
[
1792281600000,
11.42
],
[
1792368000000,
41.47
],
[
1792454400000,
95.76
],
[
1792540800000,
66.89
],
[
1792627200000,
35.24
],
[
1792713600000,
50.51
]
]
},
"reservedHours": {
"query": "sum:aws.compute.gpu_hours{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100,pricing:reserved}.rollup(sum, 86400)",
"unit": "gpu_hour",
"pointlist": [
[
1782345600000,
220
],
[
1782432000000,
220
],
[
1782518400000,
220
],
[
1782604800000,
220
],
[
1782691200000,
220
],
[
1782777600000,
220
],
[
1782864000000,
220
],
[
1782950400000,
220
],
[
1783036800000,
220
],
[
1783123200000,
220
],
[
1783209600000,
220
],
[
1783296000000,
220
],
[
1783382400000,
220
],
[
1783468800000,
220
],
[
1783555200000,
220
],
[
1783641600000,
220
],
[
1783728000000,
220
],
[
1783814400000,
220
],
[
1783900800000,
220
],
[
1783987200000,
220
],
[
1784073600000,
220
],
[
1784160000000,
220
],
[
1784246400000,
220
],
[
1784332800000,
220
],
[
1784419200000,
220
],
[
1784505600000,
220
],
[
1784592000000,
220
],
[
1784678400000,
220
],
[
1784764800000,
220
],
[
1784851200000,
220
],
[
1784937600000,
220
],
[
1785024000000,
220
],
[
1785110400000,
220
],
[
1785196800000,
220
],
[
1785283200000,
220
],
[
1785369600000,
220
],
[
1785456000000,
220
],
[
1785542400000,
220
],
[
1785628800000,
220
],
[
1785715200000,
220
],
[
1785801600000,
220
],
[
1785888000000,
220
],
[
1785974400000,
220
],
[
1786060800000,
220
],
[
1786147200000,
220
],
[
1786233600000,
197.21
],
[
1786320000000,
220
],
[
1786406400000,
220
],
[
1786492800000,
220
],
[
1786579200000,
220
],
[
1786665600000,
220
],
[
1786752000000,
220
],
[
1786838400000,
220
],
[
1786924800000,
220
],
[
1787011200000,
218.97
],
[
1787097600000,
220
],
[
1787184000000,
220
],
[
1787270400000,
220
],
[
1787356800000,
205.11
],
[
1787443200000,
220
],
[
1787529600000,
220
],
[
1787616000000,
220
],
[
1787702400000,
220
],
[
1787788800000,
220
],
[
1787875200000,
220
],
[
1787961600000,
214.38
],
[
1788048000000,
220
],
[
1788134400000,
220
],
[
1788220800000,
220
],
[
1788307200000,
220
],
[
1788393600000,
220
],
[
1788480000000,
220
],
[
1788566400000,
220
],
[
1788652800000,
220
],
[
1788739200000,
220
],
[
1788825600000,
220
],
[
1788912000000,
220
],
[
1788998400000,
220
],
[
1789084800000,
220
],
[
1789171200000,
220
],
[
1789257600000,
220
],
[
1789344000000,
220
],
[
1789430400000,
220
],
[
1789516800000,
220
],
[
1789603200000,
220
],
[
1789689600000,
220
],
[
1789776000000,
217.96
],
[
1789862400000,
220
],
[
1789948800000,
220
],
[
1790035200000,
220
],
[
1790121600000,
220
],
[
1790208000000,
220
],
[
1790294400000,
220
],
[
1790380800000,
220
],
[
1790467200000,
220
],
[
1790553600000,
220
],
[
1790640000000,
220
],
[
1790726400000,
220
],
[
1790812800000,
220
],
[
1790899200000,
220
],
[
1790985600000,
220
],
[
1791072000000,
220
],
[
1791158400000,
220
],
[
1791244800000,
220
],
[
1791331200000,
220
],
[
1791417600000,
220
],
[
1791504000000,
220
],
[
1791590400000,
220
],
[
1791676800000,
220
],
[
1791763200000,
220
],
[
1791849600000,
220
],
[
1791936000000,
220
],
[
1792022400000,
220
],
[
1792108800000,
220
],
[
1792195200000,
220
],
[
1792281600000,
220
],
[
1792368000000,
220
],
[
1792454400000,
220
],
[
1792540800000,
220
],
[
1792627200000,
220
],
[
1792713600000,
220
]
]
},
"costUsd": {
"query": "sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)",
"unit": "usd",
"pointlist": [
[
1782345600000,
3962.51
],
[
1782432000000,
3936.97
],
[
1782518400000,
3127.9
],
[
1782604800000,
2957.06
],
[
1782691200000,
3190.37
],
[
1782777600000,
3675.1
],
[
1782864000000,
3057.92
],
[
1782950400000,
3154.1
],
[
1783036800000,
3490.36
],
[
1783123200000,
3104.13
],
[
1783209600000,
2814.16
],
[
1783296000000,
3343.11
],
[
1783382400000,
3168.5
],
[
1783468800000,
3641.89
],
[
1783555200000,
3214.01
],
[
1783641600000,
3413.77
],
[
1783728000000,
3146.28
],
[
1783814400000,
3158.71
],
[
1783900800000,
3205.32
],
[
1783987200000,
3126.64
],
[
1784073600000,
2884.18
],
[
1784160000000,
3133.57
],
[
1784246400000,
3467.5
],
[
1784332800000,
2531.3
],
[
1784419200000,
2945.18
],
[
1784505600000,
2809.77
],
[
1784592000000,
2748.26
],
[
1784678400000,
2114.44
],
[
1784764800000,
3051.8
],
[
1784851200000,
3057.15
],
[
1784937600000,
2602.74
],
[
1785024000000,
2618.81
],
[
1785110400000,
3763.31
],
[
1785196800000,
3466.02
],
[
1785283200000,
3154.42
],
[
1785369600000,
2634.67
],
[
1785456000000,
2428.52
],
[
1785542400000,
2294.41
],
[
1785628800000,
2311.53
],
[
1785715200000,
2517.38
],
[
1785801600000,
2648.13
],
[
1785888000000,
2857.04
],
[
1785974400000,
3755.45
],
[
1786060800000,
2776.01
],
[
1786147200000,
2681.1
],
[
1786233600000,
1908.2
],
[
1786320000000,
2507.68
],
[
1786406400000,
2635.68
],
[
1786492800000,
3445.2
],
[
1786579200000,
2762.81
],
[
1786665600000,
2960.86
],
[
1786752000000,
2370.17
],
[
1786838400000,
2352.87
],
[
1786924800000,
3002.27
],
[
1787011200000,
2240.18
],
[
1787097600000,
2957.05
],
[
1787184000000,
2680.54
],
[
1787270400000,
2850.94
],
[
1787356800000,
2082.92
],
[
1787443200000,
2547.96
],
[
1787529600000,
3342.37
],
[
1787616000000,
3191.1
],
[
1787702400000,
3117.97
],
[
1787788800000,
3115.3
],
[
1787875200000,
3507.03
],
[
1787961600000,
2216.12
],
[
1788048000000,
2787.39
],
[
1788134400000,
3214.84
],
[
1788220800000,
3168.71
],
[
1788307200000,
3114.7
],
[
1788393600000,
2941.45
],
[
1788480000000,
3729.93
],
[
1788566400000,
2544.57
],
[
1788652800000,
2997.28
],
[
1788739200000,
3668.84
],
[
1788825600000,
3610.82
],
[
1788912000000,
2936.39
],
[
1788998400000,
2580.66
],
[
1789084800000,
3056.38
],
[
1789171200000,
2554.74
],
[
1789257600000,
2767.35
],
[
1789344000000,
2741.02
],
[
1789430400000,
3824.64
],
[
1789516800000,
3333.11
],
[
1789603200000,
3568.44
],
[
1789689600000,
3278.7
],
[
1789776000000,
2252.61
],
[
1789862400000,
2721.57
],
[
1789948800000,
3391.52
],
[
1790035200000,
3267.05
],
[
1790121600000,
3223.32
],
[
1790208000000,
3808.99
],
[
1790294400000,
3062.73
],
[
1790380800000,
2372.14
],
[
1790467200000,
3122.31
],
[
1790553600000,
3490.33
],
[
1790640000000,
2729.31
],
[
1790726400000,
3612.03
],
[
1790812800000,
2828.47
],
[
1790899200000,
3052
],
[
1790985600000,
2981.8
],
[
1791072000000,
3111.69
],
[
1791158400000,
3480.66
],
[
1791244800000,
3350.48
],
[
1791331200000,
2769.41
],
[
1791417600000,
3428.68
],
[
1791504000000,
3743.69
],
[
1791590400000,
2493.4
],
[
1791676800000,
2955.11
],
[
1791763200000,
3245.42
],
[
1791849600000,
2582.81
],
[
1791936000000,
2977.1
],
[
1792022400000,
3317.26
],
[
1792108800000,
2990.87
],
[
1792195200000,
2892.93
],
[
1792281600000,
2627.76
],
[
1792368000000,
3187.54
],
[
1792454400000,
3827.65
],
[
1792540800000,
3571.04
],
[
1792627200000,
2916.41
],
[
1792713600000,
3301.46
]
]
}
}
}
}POST /api/recommend (60% reserved, 6mo, market scenario)
{
"status": 200,
"body": {
"persisted": false,
"workload": {
"id": "11111111-1111-1111-1111-111111111111",
"name": "inference-prod-h100",
"gpuType": "H100"
},
"inputs": {
"reservedPct": 0.6,
"horizonMonths": 6,
"usageBias": 0,
"scenario": "market",
"gpuType": "H100"
},
"recommendation": {
"reservedPct": 0.6,
"horizonMonths": 6,
"gpuType": "H100",
"mix": [
{
"monthOffset": 0,
"reservedHours": 7428,
"onDemandHours": 512.63,
"costUsd": 18095.66
},
{
"monthOffset": 1,
"reservedHours": 7428,
"onDemandHours": 512.63,
"costUsd": 18135.23
},
{
"monthOffset": 2,
"reservedHours": 7428,
"onDemandHours": 512.63,
"costUsd": 18083.1
},
{
"monthOffset": 3,
"reservedHours": 7428,
"onDemandHours": 512.63,
"costUsd": 18033.68
},
{
"monthOffset": 4,
"reservedHours": 7428,
"onDemandHours": 512.63,
"costUsd": 17994.93
},
{
"monthOffset": 5,
"reservedHours": 7428,
"onDemandHours": 512.63,
"costUsd": 17917.57
}
],
"baselineCostUsd": 108260.17,
"strategyCostUsd": 108260.17
},
"savingEstimateUsd": 0,
"confidence": {
"low": 0,
"high": 0
},
"rationale": "Under the market scenario over 6 months, reserving now beats staying all-on-demand. Steady floor is 248 GPU-h/day, mean demand 265 GPU-h/day; reserving 60% of mean demand locks in $2.272/GPU-h vs a curve mean of $2.233/GPU-h — today's reservation rate is near or above the forward-curve mean.",
"breakEvenMonthOffset": 0,
"provenance": {
"ornn": {
"source": "ornn_fixture",
"fetchedAt": "2026-10-23T00:18:27.450Z",
"horizonDays": 180,
"pointCount": 180
},
"datadog": {
"from": "2026-06-25",
"to": "2026-10-23",
"queries": [
"sum:system.gpu.hours{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)",
"avg:dcgm.gpu.util{workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100} by {host}",
"sum:aws.cost.amortized{service:ec2,workload:inference-prod,env:prod,team:ml-serving,gpu_type:H100}.rollup(sum, 86400)"
]
},
"engine": {
"curvePointCount": 180,
"usageDayCount": 121,
"scenario": "market",
"horizonMonths": 6,
"reservedPct": 0.6
}
}
}
}POST /api/backtest (same strategy, 12mo window default)
{
"status": 200,
"body": {
"persisted": false,
"workload": {
"id": "11111111-1111-1111-1111-111111111111",
"name": "inference-prod-h100",
"gpuType": "H100"
},
"strategy": {
"reservedPct": 0.6,
"horizonMonths": 6,
"scenario": "market",
"usageBias": 0
},
"windowStart": "2025-10-23",
"windowEnd": "2026-10-23",
"cumulativeSavingUsd": 32902.38,
"hitRate": 1,
"worstMonth": {
"month": "2025-10",
"delta_usd": 1051.1
},
"monthly": [
{
"month": "2025-10",
"baseline_usd": 5864.56,
"strategy_usd": 4813.46,
"delta_usd": 1051.1
},
{
"month": "2025-11",
"baseline_usd": 18980.77,
"strategy_usd": 15651.26,
"delta_usd": 3329.51
},
{
"month": "2025-12",
"baseline_usd": 20192.15,
"strategy_usd": 16776.38,
"delta_usd": 3415.78
},
{
"month": "2026-01",
"baseline_usd": 22073.26,
"strategy_usd": 18479.26,
"delta_usd": 3594.01
},
{
"month": "2026-02",
"baseline_usd": 20116.19,
"strategy_usd": 16967.7,
"delta_usd": 3148.49
},
{
"month": "2026-03",
"baseline_usd": 22622.84,
"strategy_usd": 19229.22,
"delta_usd": 3393.62
},
{
"month": "2026-04",
"baseline_usd": 20716.5,
"strategy_usd": 17751.38,
"delta_usd": 2965.12
},
{
"month": "2026-05",
"baseline_usd": 20375.95,
"strategy_usd": 17604.11,
"delta_usd": 2771.84
},
{
"month": "2026-06",
"baseline_usd": 18301.91,
"strategy_usd": 15945.89,
"delta_usd": 2356.02
},
{
"month": "2026-07",
"baseline_usd": 17727.11,
"strategy_usd": 15579.68,
"delta_usd": 2147.43
},
{
"month": "2026-08",
"baseline_usd": 16859.79,
"strategy_usd": 14952.8,
"delta_usd": 1907
},
{
"month": "2026-09",
"baseline_usd": 15911.06,
"strategy_usd": 14240.14,
"delta_usd": 1670.92
},
{
"month": "2026-10",
"baseline_usd": 11773.59,
"strategy_usd": 10622.04,
"delta_usd": 1151.55
}
]
}
}