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 | 0.21× | 229 / 181 | 131 / 150 | 2 / 6 |
| 2026-08-22 | 0.20× | 244 / 453 | 135 / 393 | 2 / 6 |
| 2026-08-23 | 0.34× | 88 / 192 | 85 / 187 | 2 / 6 |
| 2026-08-24 | 0.35× | 101 / 124 | 99 / 119 | 2 / 4 |
| 2026-08-26 | 0.35× | 194 / 734 | 182 / 625 | 2 / 6 |
| 2026-08-27 | 0.41× | 186 / 429 | 175 / 322 | 2 / 6 |
| 2026-08-31 | 0.44× | 47 / 48 | 47 / 47 | 2 / 5 |
| 2026-09-03 | 0.84× | 180 / 57 | 179 / 57 | 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.