RTX 4090 vs A100
Datasheet specs side by side, plus what each card actually costs to rent today across every provider we track.
RTX 4090
A100
Spec sheet
| RTX 4090 | A100 | |
|---|---|---|
| Architecture | Ada Lovelace | Ampere |
| VRAM (max variant) | 24 GB | 80 GB |
| FP16 TFLOPS (dense) | 165 | 312 |
| Memory bandwidth | 1008 GB/s | 1935 GB/s |
| TDP | 450 W | 300 W |
| Interconnect | PCIe Gen4 | PCIe Gen4 |
| Launched | 2022 | 2020 |
Computed price-performance at today's cheapest on-demand rates: RTX 4090 $2.12 per PFLOP-hour vs A100 $2.85. Formula: cheapest $/hr ÷ dense FP16 PFLOPS (1 PFLOP = 1000 TFLOPS). Specs from vendor datasheets (source, source).
What they cost to rent (2026-08-22)
| RTX 4090 | A100 | |
|---|---|---|
| Cheapest on-demand $/hr | $0.35 | $0.89 |
| Cheapest spot/community $/hr | $0.16 | $0.68 |
| Cheapest source | TensorDock | Jarvislabs |
| Providers offering it | 4 | 21 |
Cheapest verified on-demand rate per side from today's snapshot; highlighted = cheaper. Spot/marketplace tiers shown on the individual pages.
FAQ
RTX 4090 is cheaper: the RTX 4090 rents from $0.35 per GPU-hour and the A100 from $0.89 on-demand, verified 2026-08-22.
The RTX 4090 has 24 GB and the A100 has 80 GB. The A100 has more.
Prices render from the live dataset each build. More: the screener · price movers · methodology.