← Back to the run console
completed
cp-1
cp-analysis · jump-sample-BR00116991
What ran
- Pipeline
- cp-analysis
- Dataset
- jump-sample-BR00116991
- Worker node
- ws3
- Submitted by
- you · 9h 13m ago
- Images processed
- 32 / 32
- Measurements per image
- 116
32 images in 1m11s on ws3
Run receipt
$0.0090
$0.0090 ÷ 32 images = $0.28 per 1,000 images
- Runtime
- 1m 11s
- Throughput
- 1,633 imgs/hr
- Bytes in
- 79.4 MB
| Line item | Basis | Cost |
|---|---|---|
|
Electricity Wall power on ws3 during the run |
1m 11s × 250 W × $0.13/kWh | $0.0006 |
|
Hardware (amortised) Replacement-cost depreciation over 5 years |
1m 11s × ($3,000 / 5 yr) | $0.0013 |
|
Bandwidth (AWS egress equivalent) What AWS would charge to pull these inputs out of S3 |
79.4 MB × $0.09 / GB | $0.0070 |
| ToxIndex Cell total | $0.0090 | |
|
AWS Batch equivalent c5.4xlarge on-demand × 1.10 Batch overhead + same S3 egress |
1m 11s × $0.68/hr × 1.10 + 79.4 MB × $0.09/GB | $0.02 |
| You save vs AWS | $0.01 (59% cheaper) | |
Sources
- Electricity: EIA Virginia commercial average ($0.13 / kWh) — source
- Hardware: Replacement-cost depreciation, $3,000 over 5 years (i9-7940X / 64GB / RTX 4080 build)
- Power draw: i9-7940X 165W TDP + board/RAM/drives + PSU loss → ~250W wall
- Bandwidth: AWS S3 → Internet first-10TB tier ($0.09 / GB) — source
- AWS Batch: c5.4xlarge (16 vCPU, 32 GB): $0.68/hr on-demand + 10% Batch overhead — source
Computed 2026-05-24 21:53 UTC from the run's stored runtime and input byte count. Rate-card changes only affect new runs.