12 games at 1440p, 28 FPS on average, anchored estimate · 1080p gaming, 1440p is a stretch outside older titles
28 Avg FPS Includes estimates
We have not put this card on our bench yet. The numbers below are interpolated from the cards we did measure, on the same suite and the same settings. Across our 12-game suite at 1440p, NVIDIA GeForce GTX 1660 averages 28 FPS. The ceiling is F1 2022 at 50 FPS; the floor is The Last of Us Part I at 16 FPS, a 3.1x spread that tells you how engine-dependent this class of card is. In practical terms that means 1080p gaming, 1440p is a stretch outside older titles. Under load we logged a peak of 120W against the 120W rating.
| Benchmark | Result | 1% Low | VRAM Used | Power | Data |
|---|---|---|---|---|---|
| Atomic Heart 1440p · Atomic | 39 FPS | 29 | 5.33 GB | 97 W | Estimated |
| Call of Duty: MW2 1440p · Extreme | 24 FPS | 17 | 5.12 GB | 85 W | Estimated |
| Cyberpunk 2077 1440p · Ultra | 25 FPS | 20 | 4.96 GB | 92 W | Estimated |
| Dying Light 2 1440p · High Quality | 28 FPS | 23 | 3.77 GB | 84 W | Estimated |
| Elden Ring 1440p · Maximum | 30 FPS | 26 | 3.93 GB | 120 W | Estimated |
| F1 2022 1440p · Ultra High | 50 FPS | 24 | 5.06 GB | 91 W | Estimated |
| Forza Horizon 5 1440p · Extreme | 30 FPS | 26 | 5.58 GB | 120 W | Estimated |
| Hogwarts Legacy 1440p · Ultra | 23 FPS | 12 | 5.18 GB | 90 W | Estimated |
| Red Dead Redemption 2 1440p · Favor Quality | 29 FPS | 23 | 5.15 GB | 100 W | Estimated |
| Resident Evil 4 Remake 1440p · Maximum | 29 FPS | 22 | 5.31 GB | 101 W | Estimated |
| Star Wars Jedi: Survivor 1440p · Epic | 18 FPS | 11 | 5.33 GB | 92 W | Estimated |
| The Last of Us Part I 1440p · Ultra | 16 FPS | 13 | 5.58 GB | 78 W | Estimated |
| Architecture | Turing (TU116) |
| CUDA cores | 1,408 |
| VRAM | 6GB GDDR5 |
| Memory bus | 192-bit |
| Memory bandwidth | 192 GB/s |
| Boost clock | 1,785 MHz |
| TDP | 120 W |
| Process | 12nm |
| Interface | PCIe 3.0 x16 |
| Release date | 2019-03-14 |
| Launch MSRP | $219 |
NVIDIA GeForce GTX 1660 averages 28 FPS at 1440p on our suite. We have not put this card on our bench yet. The numbers below are interpolated from the cards we did measure, on the same suite and the same settings.
100% = this card, Gaming headline metric (Avg FPS). #75 of 82 desktop cards in this vertical.
| GPU | Relative | % | Avg FPS |
|---|---|---|---|
| NVIDIA GeForce GTX 1660 Super | 111% | 31 | |
| AMD Radeon RX 6500 XT | 107% | 30 | |
| NVIDIA GeForce GTX 1650 Super | 104% | 29 | |
| NVIDIA GeForce GTX 1060 6GB | 100% | 28 | |
| NVIDIA GeForce GTX 1660 | 100% | 28 | |
| NVIDIA GeForce GTX 980 | 93% | 26 | |
| Intel Arc A380 | 86% | 24 | |
| NVIDIA GeForce GTX 780 Ti | 82% | 23 | |
| AMD Radeon RX 6400 | 79% | 22 |
Same card, other workloads: NVIDIA GeForce GTX 1660 AI benchmarks, LLM tokens/sec, image generation & VRAM fit
| Transistors | 6,600 million |
| Die size | 284 mm² |
| Process node | 12 nm |
| Fabricated by | TSMC |
| Transistor density | 23.2 million per mm² |
Denser than 64% of the 746 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.
Can't run: 3D game asset kit (needs Stable Diffusion XL), Product shoot, start to finish (needs Stable Diffusion XL), Product photo shoot (needs Stable Diffusion XL), Photo restoration batch (needs FLUX.1 Kontext dev), Restore and enlarge photos (needs FLUX.1 Kontext dev), Character sheet, 12 poses (needs FLUX.1 dev), Short social clips (needs Qwen3 32B), 60-second AI short film (needs Qwen3 32B), 24-frame storyboard (needs Qwen3 32B), 6-panel comic page (needs Qwen3 32B), Long-form article batch (needs Llama 3.3 70B), Full codebase review (needs Qwen2.5-Coder 14B).
This card is $219 to buy. The cheapest listed rate on Vast.ai is $0.049/hour, but that is the floor: we budget $0.059/hour, a 20% premium, because idle time, storage and unavailable cheap instances all land on the same bill. At that rate buying wins after 3,724 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 | $43 | 5.1 years |
| 8 hours a day, working on it | 2,920 | $172 | 1.3 years |
| 24/7, always-on agent | 8,760 | $515 | 5.1 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.