A paid production workload on the reference fleet, measured on GeForce RTX 4080 and GeForce RTX 5090 on the same days. One workload, its own cells — never averaged with any other workload, because cross-workload aggregation is the substitution error this project exists to expose.
| DAY | GOODPUT GeForce RTX 4080 / GeForce RTX 5090 | ATTEMPTS | COMPLETIONS | CARDS |
|---|---|---|---|---|
| 2026-08-21 | 0.51× | 12 / 15 | 11 / 15 | 2 / 4 |
| 2026-08-22 | 0.51× | 12 / 15 | 11 / 15 | 2 / 4 |
| 2026-08-23 | 0.45× | 39 / 33 | 37 / 33 | 2 / 6 |
| 2026-08-24 | 0.51× | 40 / 22 | 38 / 22 | 2 / 6 |
| 2026-08-26 | 0.46× | 44 / 137 | 41 / 132 | 2 / 6 |
| 2026-08-27 | 0.70× | 12 / 43 | 11 / 36 | 2 / 5 |
| 2026-08-29 | 0.37× | 32 / 24 | 32 / 24 | 2 / 4 |
| 2026-09-03 | 0.67× | 12 / 18 | 11 / 18 | 2 / 4 |
The measured pair page for these deployments: side-by-side cells. Downloads: ratio cells, value board, the rental prices these were costed at.