9 games at 1440p, 104 FPS on average, anchored estimate · smooth 100+ FPS 1440p gaming across demanding AAA titles
104 Avg FPS Includes estimates
This card is still waiting on bench time. Until it gets there the figures are anchored estimates, interpolated from cards we measured on the same suite. Across our 9-game suite at 1440p, NVIDIA GeForce RTX 4070 averages 104 FPS. The ceiling is Counter-Strike 2 at 168 FPS; the floor is Alan Wake 2 at 62 FPS, a 2.7x spread that tells you how engine-dependent this class of card is. In practical terms that means smooth 100+ FPS 1440p gaming across demanding AAA titles. Under load we logged a peak of 211W against the 200W rating.
Anchored estimate, this card hasn't been through our test bench yet. Its numbers are interpolated from the cards that have, on the same suite and the same settings, and they'll be replaced with measured results once it runs.
| Benchmark | Result | 1% Low | VRAM Used | Power | Data |
|---|---|---|---|---|---|
| Alan Wake 2 1440p · Ultra | 62 FPS | 53 | 9.03 GB | 191 W | Estimated |
| Baldur's Gate 3 1440p · Ultra | 120 FPS | 98 | 7.26 GB | 211 W | Estimated |
| Black Desert Online 1440p · Ultra | 125 FPS | 91 | 4.77 GB | 209 W | Estimated |
| Counter-Strike 2 1440p · Ultra | 168 FPS | 91 | 5.94 GB | 172 W | Estimated |
| Cyberpunk 2077 1440p · Ultra | 78 FPS | 62 | 6.42 GB | 183 W | Estimated |
| Diablo 4 1440p · Ultra | 101 FPS | 78 | 9.66 GB | 183 W | Estimated |
| Escape from Tarkov 1440p · Ultra | 118 FPS | 91 | 10.51 GB | 196 W | Estimated |
| Red Dead Redemption 2 1440p · Ultra | 85 FPS | 73 | 8.85 GB | 185 W | Estimated |
| The Witcher 3 1440p · Ultra | 81 FPS | 66 | 5.5 GB | 193 W | Estimated |
| Architecture | Ada Lovelace |
| CUDA cores | 5,888 |
| VRAM | 12GB GDDR6X |
| Memory bus | 192-bit |
| Memory bandwidth | 504 GB/s |
| Boost clock | 2,475 MHz |
| TDP | 200 W |
| Process | 4nm (TSMC 4N) |
| Interface | PCIe 4.0 x16 |
| Release date | 2023-04-13 |
| Launch MSRP | $599 |
NVIDIA GeForce RTX 4070 averages 104 FPS at 1440p on our suite. This card is still waiting on bench time. Until it gets there the figures are anchored estimates, interpolated from cards we measured on the same suite.
100% = this card, Gaming headline metric (Avg FPS). #25 of 82 desktop cards in this vertical.
| GPU | Relative | % | Avg FPS |
|---|---|---|---|
| AMD Radeon RX 7800 XT | 113% | 118 | |
| AMD Radeon RX 9070 GRE | 110% | 114 | |
| GeForce RTX 5070 | 108% | 112 | |
| NVIDIA GeForce RTX 3070 Ti | 106% | 110 | |
| NVIDIA GeForce RTX 4070 | 100% | 104 | |
| NVIDIA Titan RTX | 100% | 104 | |
| NVIDIA GeForce RTX 2080 Ti Founders Edition | 96% | 100 | |
| NVIDIA GeForce RTX 3080 | 96% | 100 | |
| AMD Radeon RX 6800 XT | 95% | 99 |
Same card, other workloads: NVIDIA GeForce RTX 4070 AI benchmarks, LLM tokens/sec, image generation & VRAM fit
| Transistors | 35,800 million |
| Die size | 294.5 mm² |
| Process node | 4 nm |
| Fabricated by | TSMC |
| Transistor density | 121.6 million per mm² |
Denser than 91% of the 76 cards we have silicon data for. Density is the clearest measure of what a process node bought: a card that gained it without growing the die got its speed from the fab rather than the architecture.
Silicon figures from Wikipedia (CC BY-SA 4.0). Benchmarks on this page are our own. Compare every chip.
Whole-job timings, composed from our measured per-model results on this card.
| Workflow | Time | Energy | Basis |
|---|---|---|---|
| Full codebase review | 19.7 min | 43.13 Wh | measured |
Can't run: 60-second AI short film (needs Qwen3 32B), 60-second AI short film, narrated (needs Qwen3 32B), 10 short social clips (needs Qwen3 32B), 6-panel comic page (needs Qwen3 32B), 20 long-form articles (needs Llama 3.3 70B), Character sheet, 12 poses (needs FLUX.1 dev), 100-photo restoration batch (needs FLUX.1 Kontext dev), 24-frame storyboard (needs Qwen3 32B), 100-photo restore and enlarge (needs FLUX.1 Kontext dev).
This card is $599 to buy. The cheapest listed rate on Vast.ai is $0.076/hour, but that is the floor: we budget $0.091/hour, a 20% premium, because idle time, storage and unavailable cheap instances all land on the same bill. At that rate buying wins after 6,568 GPU-hours. Below it you are paying for idle silicon.
| How you would use it | GPU-hours a year | Rental cost a year | Time to break even |
|---|---|---|---|
| 2 hours a day, hobby | 730 | $67 | 9.0 years |
| 8 hours a day, working on it | 2,920 | $266 | 2.2 years |
| 24/7, always-on agent | 8,760 | $799 | 9.0 months |
At hobby usage this card is very unlikely to pay for itself before it is superseded. Rent it. Rental figures include a 20% premium over the cheapest listed rate. Ignores electricity, resale and the fact that a rented card can be a newer one tomorrow.
Cheapest of the RunPod and Vast on-demand rates we see, sampled daily. Spot and interruptible pricing runs lower.