Thief
This combination provides smooth gameplay with an average of 60 FPS, suitable for most gaming scenarios.
Average FPS: resolution vs quality settings
| Ultra | High | Medium | Low | |
|---|---|---|---|---|
| 4K Ultra HD | 27 | 28 | 37 | 48 |
| 1440p QHD | 45 | 47 | 62 | 82 |
| 1080p Full HD | 66 | 69 | 91 | 120 |
Cell color: green is 120+ FPS, teal is 60+, amber is 30+, red is below 30.
Thief with Intel Core i5-14600KF + NVIDIA GeForce GTX 1660 SUPER, In-Depth Analysis
The Intel Core i5-14600KF pairing with the NVIDIA GeForce GTX 1660 SUPER produces a combination where the CPU’s top-tier compute performance (90th percentile) is far ahead of the GPU’s mid-tier 3D capability (53rd percentile). For the 2014 stealth title Thief, the measured frame rates at 1080p, 1440p, and 4K consistently show that the battlefield is limited by the graphics card, not the processor. At every resolution, the CPU is capable of feeding the display at 120 FPS on Low settings, but the frame rate collapses as resolution climbs, reflecting a GPU bottleneck. The overall combination ranks 3179 of 3720 tested combos, confirming that while the CPU is elite among processors, this particular match is held back by the GPU when the game pushes pixel work.
Resolution Scaling
At 1080p, the Low preset delivers an average of 120 FPS, whereas High lands at 69 FPS and Ultra at 66 FPS. This shows the GPU can saturate a 1080p display at low visual load. Moving up to 2560×1440, Low drops to 82 FPS, High to 47 FPS, and Ultra to 45 FPS. At 4K (3840×2160), the numbers tumble further: Low averages 48 FPS, High falls to 28 FPS, and Ultra bottomts out at 27 FPS. The trend is clear: the GPU becomes the single limiting factor when resolution increases, because the CPU headroom remains large — even at 4K Low it manages 48 FPS, but at 4K Ultra, it sinks to 27 FPS due to pixel load.
The differential between Low and Ultra also squeezes as resolution rises: at 1080p the gap is 54 FPS (120 vs 66), at 1440p it is 37 FPS (82 vs 45), and at 4K it is only 21 FPS (48 vs 27). This decreasing gap indicates that the GPU’s processing capacity is saturated earlier at higher resolutions, so settings differences become less impactful because the renderer is already strained. In practical terms, a player who steps down from Extreme to Low sees a massive improvement at 1080p but only a 18 FPS gain at 4K, implying that the GPU is the real wall and the Mi resolution too heavy for the card.
How This Combo Ranks
In the benchmark database, the Core i5-14600KF + GTX 1660 SUPER lands at position 3179 out of 3720 combinations, which means it performs worse than about 85% of the tested pairs. This rank is surprising given the CPU’s overall position — the i5-14600KF, with its 14 cores and 20 threads, sits in the 90th percentile across all CPUs, near the AMD Ryzen 9 7900 (0.3% ahead), Ryzen 7 PRO 5755G (0.4% ahead), and Intel Core Ultra 6 but 245 (0.8% ahead), while trailing only the AMD Ryzen AI Max+ 388 (-0.8%). But the GPU’s percentile is only 53, with its closest rivals including 3050 Ti Mobile (0.4% ahead) and RX 580 (0.4% ahead). In game-specific benchmarks, the GPU’s PassMark G3D score is 12699, which it also sits close to these other cards. The rank of 3179 indicates that the GPU’s poor scaling cannot leverage the processor’s massive compute power in Thief, since the game engine is single-thread friendly and the GPU is the true constraint.
GPU Role
The GTX 1660 SUPER, built on the Turing architecture with a TU116 chip, features 1408 shading units, 88 TMUL, and 48 ROP. Its base clock is 1530 MHz with a boost clock of 1785 MHz. Memory is 6 GB of GDDR6 on a 192-bit bus, delivering 336 GB/s of bandwidth with an effective memory clock of 14 Gbps. These numbers explain the FPS pattern: the modest texture memory and a mid-range compute keep the card effective for 1080p but not for high-ratio downscaling. The measured data shove that average FPS at 4K Ultra is 27, and the minimum at 1440p Medium is 53, showing that the GA boundary lines are hard low.
The GPU has a PassMark G3D score of 12699, placing it 53rd percentile among all GPUs. In terms of performance range, it sits between the RTX Ti 3050 Ti (0.4% higher) and RX 580 (0.4%) at lower end, and is 0.9% above the actual AMD Radeon 760M. Its pre-measured direct 3D numbers are consistent: PassMark Direct 2D 10 = 65, Direct 3D = 104, DirectSpec = 50. In-game, the GPU’s power is the deciding factor. The 6 GB VRAM is not exceeded, but the 2160p Ultra setting doesn’t reach a playable 30 FPS, pulling only 27 FPS, which is better suited for 1080p. The card’s GPU compute force is 5076 in PassMark (GPU Compute), and this aligns with the observed framerate scaling.
The memory bandwidth of 336 GB/s and 192-bit bus is fit for 1080p rendering; at 1440p and 4K, the pixel-throughput drops because more pixels per frame must be shaded. The card’s pixel rate is 85.68 GPixel/s and texture rate 157.1 GTexel/s, but these are theoretical. At 4K Ultra the 27 FPS average indicates that fillrate and shading are saturated. Meanwhile, it still can manage 82 FPS at 1440p Low, proving the GPU is above a 1080p baseline when the scene is light.
FAQ
*Q: Is the Intel Core i5-14600KF a good match for the GTX 1660 SUPER in Thief?*
A: The data shows a clear CPU bottleneck is absent; the i5-14600KF can deliver 120 FPS at 1080p Low, but the GPU forces it down to 66 FPS at the same resolution on Ultra. So the CPU far exceeds the GPU’s ceiling.
Q: What is the highest preset that keeps average FPS above 60 at 1440p?
A: At 2560×1440, only the Low and Medium presets stay above 60 FPS: Low averages 82 FPS, and Medium is 62 FPS. High and Ultra drop to 47 and 45 FPS, respectively.
Q: Can this setup run at 60 FPS on Ultra at 4K?
A: No. Even the Ultra preset at 3840×2160 reaches only 27 FPS on average. Low hits 48 FPS, still under a 60 FPS target.
Q: How does the GPU compare to its nearest rivals based on the database?
A: In the GPU database, the GTX 1660 SUPER is 0.4% slower than an RTX 3050 Ti Mobile and a RX 580, and 0.1% faster than a Radeon 760M. This places it in a common mid-range band.
Q: Is the combo rank impacted by the CPU’s high percentile?
A: The CPU is in the 90th percentile, yet the combination ranks 3179 of 3720. The low rank is caused primarily by the GPU, which is in the 53th percentile and thus barely above average in compute.
Q: Does resolution affect the min of frame rate for Medium settings?
A: Yes. For Medium presets, the measured average and min/max numbers are: 1080p: avg 91, min 77, max 105; 1440p: avg 62, min 53, max 72; 4K: avg 37, min 31, max 42.
CPU Role
The Intel Core i5-14600KF is based on Raptor Lake (Raptor Lake-R) and has 14 cores and 20 threads, running at a base 3.50 GHz and a boost of 5.30 GHz. It carries a 24 MB of L3 cache shared across the silicon, along with 2 MB of L2 per core. In synthetic tests, it scores 32544 in Cinebench R23 multicore and 4594 in single‑core, and hits 17085 in Geekbench multicore. These scores, along with a PassMark multithread of 38698, place it in the 90th percentile among desktop processors.
Yet in Thief, the visible frame rates are limited by the GPU, not the CPU capability. The processor easily drives the game for running at 1080p up to 120 FPS (Low settings), but the library limit stops the frame rate at higher resolutions. The preserving headroom implies that the i5-14600KF would appear identical if paired with a more powerful graphics card, while the combination gets its rank from the GPU. On the nearestRivals list, the CPU outperforms or matches the AMD Ryzen 9 7900 (avgScore 49228 vs 223) by a −0.3%, and beats the Ryzen 7 780G by 0.4% – all small differences that serve as evidence that the CPU is not the limiting factor.
For example, at 1080p the CPU can push 120 FPS, but Ultra settings only on 66 FPS – a huge gap that is entirely attributable to the GPU. Even at 4K the CPU remains idle, since the average 27 FPS at 4K Ultra is ten times lower than the processor’s parallel performance ceiling. The game is 14 years old and likely utilises that many threads, but the processor’s multiple cores are unused. Thus, in this benchmark, the CPU is premium but not the determining component.
Settings Recommendations
Based on the measured data, the Medium preset appears to give the sweet spot for both 1080p and 1440p. At 1080p, Medium maintains an average of 91 FPS with a minimum of 77, ensuring smooth gameplay on typical monitors. The drop to High only improves the minimum but sacrifices about 22 FPS, because the average slips to 69 FPS. At 1440p, Medium is still playable at an average 62 FPS with a minimum of 53, which remains above the 60 FPS floor. High retains a stable bump at 47 FPS, which is marginally playable, but Ultra drops to 45 FPS, risking stutter in more crowded scenes.
Low is excessive for 1080p users: 120 FPS is great on 144 Hz displays, but if the game’s visual fidelity is needed, Medium offers a favourable balance. On 4K, Medium is the only setting that achieves an average higher than 30 FPS (37 FPS), but with a minimum of 31 FPS it may show occasional stutter. For those who prefer Lower latency and a higher frame-time margin, the Low preset at 4K reaches 48 FPS, but this requires sacrificing most visual effects. Given that Ultra provides only 27 FPS at 4K, it is not recommended for serious play. In summary, the data suggests Medium is the best option for consistent frame pace at 1080p and 1440p, while 4K players should stay on Low to maintain a cross‑stable 48 FPS with the option to use Medium if visuals are prioritized.
Measured FPS Breakdown
Here is the full record from the database, showing average (and min/max, when captured) for each resolution and preset:
1920×1080
- Low: 120 FPS
- Medium: average 91, min 77, max 105
- High: 69 FPS
- Ultra: 66 FPS
2560×1440
- Low: 82 FPS
- Medium: average 62, min 53, max 72
- High: 47 FPS
- Ultra: 45 FPS
3840×2160
- Low: 48 FPS
- Medium: average 37, min 31, max 42
- High: 28 FPS
- Ultra: 27 FPS
The minimum frames, which are provided only for Medium, give a glimpse of the worst-case frame delivery: at 1080p it is 77, at 1440p it is 53, and at 4K it is 31. These min/max data underline the GPU degradation as resolution rises. For example, the range at 4K Medium – from 31 to 42 – shows stutter and frame pacing dips that approach the low end of playability. In contrast, 1080p Medium has a smooth min of 77, which is high enough to avoid hitches. The averages themselves are sufficient to judge that the card’s 6 GB VRAM is not overflow, but the rendering load over the 4K screen is the constraint.
All frames in the database are the measured output of the i5-14600KF and GTX 1660 SUPER configuration on Thief. The table confirms the dominance of the GPU: even at High settings, the frame rate halves from 1080p to 4K (from 69 to 28), while the CPU’s borrowed power never changes.
Hardware Specifications
Intel Core i5-14600KF
NVIDIA GeForce GTX 1660 SUPER
FPS Benchmarks by Resolution & Settings
Performance data for Intel Core i5-14600KF + NVIDIA GeForce GTX 1660 SUPER in Thief
| Resolution | Settings | Avg FPS |
|---|---|---|
| 3840x2160 | Low | 48.0 |
| 3840x2160 | Medium | 37.0 |
| 3840x2160 | High | 28.0 |
| 3840x2160 | Ultra | 27.0 |
| 2560x1440 | Low | 82.0 |
| 2560x1440 | Medium | 62.0 |
| 2560x1440 | High | 47.0 |
| 2560x1440 | Ultra | 45.0 |
| 1920x1080 | Low | 120.0 |
| 1920x1080 | Medium | 91.0 |
| 1920x1080 | High | 69.0 |
| 1920x1080 | Ultra | 66.0 |
Thief with ii5-14600KF + Other GPUs
See how Thief performs with the same CPU but different graphics cards.
Thief with GTX 1660 SUPER + Other CPUs
See how Thief performs with the same GPU but different processors.
Other Games with ii5-14600KF + GTX 1660 SUPER
Explore FPS benchmarks for other games using this same hardware combination.