Thief

The pairing of the Intel Core i7-12700K and the AMD Radeon RX 6600 delivers measurable frame rates in Thief, a 2014 stealth title whose performance profile is shaped almost entirely by resolution and preset selection. The data here is measured, not projected, and every figure below is drawn directly from the measurement set. The i7-12700K is a 12-core, 20-thread processor with a base clock of 3.60 GHz and a boost clock of 5.00 GHz, built on the 10 nm Alder Lake architecture with 25 MB of shared L3 cache. The RX 6600 is an 8 GB GDDR6 card with a 128-bit bus and 224.0 GB/s bandwidth, and its boost clock reaches 2491 MHz. Together they yield a combo rank of 2892 out of 3720 measured configurations in this title, which places them in the lower tier of tested setups. This report breaks down the measured FPS across three resolutions, explains the CPU and GPU roles, recommends settings, and answers common questions using only the data provided.

Resolution Scaling

The measured FPS at the High quality preset shows a clear and expected downward trend as resolution climbs. At 1080p, the average is 82 FPS. Moving to 1440p, average drops to 56 FPS, a reduction of 26 FPS (a 31.7% drop). At 4K, average falls further to 33 FPS, which is 41% lower than 1440p and 60% lower than 1080p. The pattern is consistent across all presets: Low starts at 142 FPS at 1080p, 98 at 1440p, and 57 at 4K. Medium averages 108, 74, and 44 across the three resolutions. Ultra averages 78, 54, and 32. The scaling is steep, typical of a GPU that is working progressively harder at higher pixel counts.

The magnitude of the scaling relative to the CPU's headroom can be inferred from the low-set figures. At 1440p with Low settings, the average is 98 FPS, which is already above what most displays can show, and at 1080p Low it hits 142 FPS. At the higher resolutions (4K), even Low stays at 57 FPS, barely above 30. This indicates that the RX 6600, not the CPU, becomes the limiting factor as resolution increases. The i7-12700K is far more capable than the FPS numbers suggest—it can feed much higher frame rates, at least at 1080p Low where the 142 FPS shows the CPU is not a bottleneck. At 4K, however, the drop to 33 FPS (High) or 32 (Ultra) shows the GPU is the constraining component.

Examining the difference between preset quality at a given resolution reveals a slight efficiency curve. At 1080p, moving from High to Ultra costs only 4 FPS (82 to 78), while High to Medium costs 26 FPS (82 to 108) in the opposite direction—that is, Medium is faster than High but High is closer to Ultra in performance. At 1440p, High is 56 FPS, Ultra is 54 (4% slower), and Medium is 74. At 4K, High is 33, Ultra is 32, and Medium is 44. The smallest gap between presets is always between High and Ultra, meaning Ultra gives almost identical framerate to High across all resolutions. Therefore, the GPU is bound in these high settings, and the resolution sensitivity is the primary factor limiting playability.

## Recommendations

Based solely on the measured FPS rows, there is no single preset that wins at every resolution. At 1080p, all four presets provide playable averages: Low 142, Medium 108, High 82, Ultra 78. The minimum recorded framerate for Medium is 92 FPS (from minFps=92), which is still smooth. The decision becomes more definite at 1440p: only Low (98) and Medium (74) are comfortably above 60 average, while High (56) and Ultra (54) are below 60, which some players will find less fluid. The 1440p Medium preset also shows min 63, max 85, so it stays within a playable envelope. At 4K, the picture changes entirely: Medium gives an average of 44 FPS with min 37 max 50, which is borderline; High and Ultra are 33 and 32, respectively; Low is 57, but even 57 is not a locked 60. For immersive 4K play, the data suggests Medium is the highest preset that keeps the average above 40 FPS, while Low is the only preset that exceeds 55 FPS at 4K.

Therefore, the optimal preset depends on the target resolution and the player's tolerance for frame variability. On 1080p, either Medium or Ultra works, but Medium's large gap to High (74 vs 56) and its min 92 / max 124 makes it the most stable relative to the real gaming experience. On 1440p, Medium (74 avg, 63 min) offers the best balance; High and Ultra sink to 56 and 54, respectively, which cannot consistently hit 60. On 4K, Medium (44 avg) is the only preset that stays above 40; Low (57) is the only one to break 50, but Low's quality penalties are not justified for a game from 2014 that can run at medium to high at 1080p. Overall, for the majority of players who choose 1440p or 1080p, Medium is the recommended preset, as it delivers the highest average FPS while still offering decent quality; at 4K, Medium is a compromise.

CPU Role

The i7-12700K is a high-tier processor with 12 physical cores and 20 threads, a base clock of 3.60 GHz, a boost clock of 5.00 GHz, and 25 MB of shared L3 cache. Its separate benchmark results place it in the 84th percentile among all CPUs, with an average benchmark score of 35287. When compared to its nearest rivals, the i7-12700KF (deltaPct -0.2) and the i7-13700T (-0.3) are essentially the same performance, while the i5-13600T (-0.1) and the Core 5 213PE (-0.4) are marginally below. This CPU has no problem driving high frame rates in any game sense from the GPU.

In Thief, the CPU’s in this combination appears to be far in advance of the 4K FPS numbers. At 1080p Low, the system hits 142 FPS, which is substantially above the 60 FPS common monitor limit. Even at 1080p Ultra, the average is 78 FPS, and the CPU is not the limiter there. The measured FPS at 1440p Low (98) and 1080p Low (142) show a large headroom from the CPU; the GPU cannot maintain that frame rate at higher resolutions or higher presets. Thus, if one were to pair this CPU with a faster graphics card, Thief would likely run at much higher FPS, but this specific GPU caps it.

The CPU role in Thief can be summarized by the low-setting numbers: the i7-12700K is far from a bottleneck. The fact that the 1080p Low average exceeds the 1080p Medium average by 34 FPS (142-108) and the 1080p High is 142, indicates that the GPU begins to throttle the pipeline as settings are pushed. When quality increases, the number of constructs drawn per frame rises, and the GPU becomes the stubborn constraint. The CPU’s core count and clock speed do not matter once the GPU is saturated; the figures confirm that no CPU-related constraint appears in the measured data.

FAQ

Q: What is the average FPS at 1080p with Medium settings?

A: The measured average is 108 FPS, with a minimum of 92 and maximum of 124.

Q: Which preset is best for a 60 FPS target at 1440p?

A: Only Low and Medium exceed 60 FPS on average; Low averages 98 FPS, and Medium 74 FPS, while High (56) and Ultra (54) fall below 60.

Q: How does the GPU’s frame rate compare from 1080p to 4K at the High preset?

A: At High, the average drops from 82 FPS at 1080p to 56 FPS at 1440p and 33 FPS at 4K—a cumulative 60% reduction from 1080p to 4K.

Q: What is the CPU’s percentile rank among all CPUs?

A: The i7-12700K sits in the 84th percentile, meaning it outperforms 84% of processors in synthetic benchmarks.

Q: At 4K with Ultra settings, what FPS can be expected?

A: The average is 32 FPS, with no minimum or maximum reported.

Q: Is 4K playable with this combo?

A: On Lower settings, the average is 57 FPS, which approaches playability, but at Medium, High, or Ultra it drops to 44, 33, and 32 respectively, making a consistent 60 FPS impossible at the higher quality presets.

GPU Role

The AMD Radeon RX 6600 is a mid-range card from the Radeon RX 6000 series, based on the Navi 23 GPU with an RDNA 2.0 architecture. It has 8 GB of GDDR6 memory on a 128-bit bus with 224.0 GB/s bandwidth, a base clock of 1626 MHz and a boost of up to 2491 MHz. The renderer features 1792 shading units, 112 TMUs, and 64 ROPs, with 28 RT cores. Its pixel rate is 159.4 GPixel/s, texture rate 279.0 GTexel/s, and FP32 performance is 8.928 TFLOPS. The GPU holds a 63rd percentile, and its average benchmark score is 19036. It closest rival, the NVIDIA Quadro K6000, is exactly equal (delta 0), while the RTX 4060 Mobile is 0.1% faster, the Tesla K20m is 0.3% faster, and the RTX 2000 Ada is 0.4% slower.

In this game, the GPU is the limiting factor at high resolutions and high presets. Evidence comes from the steep FPS drop with resolution: at 1080p High, 82 FPS; at 1440p High, 56 (32% drop); at 4K High, 33 (41% drop). The same pattern appears as preset rises: at 1080p, Low 142, Medium 108, High 82, Ultra 78. The GPU saturates before the CPU reaches its limit; the 1080p Low (142) is above the CPU’s 142 figure, while the 4K Ultra (32) brings the GPU to its knees. The 8 GB GDDR6 memory is sufficient for 1080p and 1440p textures, but at 4K the memory bandwidth may also contribute to the low frame average, as bandwidth 224 GB/s is modest.

The deltaPct against the nearest rivals confirms the card’s placement. With a 0% difference to the Quadro K6000 and -0.1% to the mobile 4050, the RX 6600 is exactly on par with a professional/workstation GPU that he lack of modern features. In Thief, these differences do not affect the present data, as the GPU is enough at 1080p and 1440p to exceed 60 for many cases, but at 4K it struggles. The fact that Ultra is only 1 FPS slower than High at 4K (32 vs 33) and 2 FPS slower at 1440p (54 vs 56) indicates that ultra settings add little stress beyond high, but the base system is already too slow.

Measured FPS Breakdown

A full table of all measured frames is provided below. The data includes each resolution and preset, with average FPS and, where available, minimum and maximum values.

• 3840x2160 (4K)
• Low: avg 57
• Medium: avg 44, min 37, max 50
• High: avg 33
• Ultra: avg 32

• 2560x1440 (1440p)
• Low: avg 98
• Medium: avg 74, min 63, max 85
• High: avg 56
• Ultra: avg 54

• 1920x1080 (1080p)
• Low: avg 142
• Medium: avg 108, min 92, max 124
• High: avg 82
• Ultra: avg 78

These numbers confirm the framerate hierarchy: Low > Medium > High > Ultra at every resolution, with the variance between High and Ultra being minimal at all three sizes (82→78, 56→54, 33→32). The minimum and maximum are only recorded for Medium, giving a picture of stability: the gap between min and max is roughly 16% to 22% of the average (e.g., 92 to 124 at 1080p, 63 to 85 at 1440p, 37 to 50 at 4K). No other presets provide min/max, so only the average is available for those initiatives.

How This Combo Ranks

The Intel i7-12700K + AMD RX 6600 combination is ranked at 2892 out of 3720 tested configurations for the game. This rank indicates that roughly 77.7% of the measured combos are faster in this particular title, placing the pairing in the lower quartile of systems. The rank is a composite of CPU, GPU, and game-specific benchmarking. Given that the CPU is in the 84th percentile and the GPU is in the 63rd percentile, the low rank is likely dominated by the GPU’s performance ceiling at high resolutions. The 1080p results are excellent, with most presets above 60 FPS, but the game’s reliance on resolution scaling pushes the average down when the display is bumped up.

Using the nearestRivals data for both CPU and GPU, the deltas show that the CPU is within 0.1% of the best near performers, while the GPU is equal to a Quadro K6000 but slower than lighter mobile cards. These hardware scores do not directly translate to a single game, but they help contextualize why this combo sits where it does. The rank certificate does not include other combos, but the position relative to 3720 total confirms that higher-end GPU selections would move this system upward, while the CPU would not limit such change. For a player working with this exact silicon, the data suggests investing in a more capable GPU would gain more FPS per dollar than changing the CPU, though no prices are discussed here. The measured FPS at 1080p Low (142) and 1440p Low (98) are already high, so the bottleneck is confirmed and the rank is exactly what the numbers imply.