Thief

# Thief with AMD Ryzen 5 7500X3D + NVIDIA GeForce RTX 5080

The measured frame rates for this combination in Thief reveal a system that is overwhelmingly GPU-limited at higher resolutions but shows signs of CPU headroom at 1080p, with the RTX 5080's raw throughput dominating the scene until pixel count drops enough to expose the Ryzen 5 7500X3D's six-core architecture. Across the twelve measured data points, the pattern is consistent: Ultra settings produce the lowest average FPS at every resolution, while Low settings produce the highest, with Medium and High falling predictably between them. The spread between Low and Ultra narrows dramatically as resolution drops, which hints at where the bottleneck shifts.

Resolution Scaling

The FPS drop from 1080p to 4K is steep but not uniform across settings, which tells a clear story about the limiting component. At Low settings, the average frame rate falls from 172 FPS at 1920x1080 to 151 FPS at 2560x1440—a 12.2% reduction—and then to 127 FPS at 3840x2160, another 15.9% drop. That total 26.2% reduction from 1080p to 4K is substantial, but it is actually the smallest relative decline among all four settings. At High settings, the averages go from 141 FPS at 1080p to 124 FPS at 1440p (12.1% drop) and then to 79 FPS at 4K (36.3% drop from 1440p), for a 44.0% total reduction. Medium settings show 155 FPS at 1080p, 137 FPS at 1440p, and 103 FPS at 4K—a total drop of 33.5%. Ultra settings show 138 FPS at 1080p, 122 FPS at 1440p, and 75 FPS at 4K, a total reduction of 45.7%.

The pattern is revealing: the higher the settings preset, the more punishing the jump to 4K becomes. Low settings lose about a quarter of their performance going from 1080p to 4K, while Ultra loses nearly half. This suggests that at 4K with Ultra settings, the GPU is being pushed to its absolute limits, and the 75 FPS average reflects that strain. Conversely, the relatively modest 26.2% drop at Low settings indicates that even at 4K, the CPU is still feeding the GPU adequately—the frame rate remains high at 127 FPS because the GPU workload is light enough that pixel throughput, not shader complexity, becomes the constraint.

The 1440p to 4K transition is the steepest segment for every settings preset. High drops 36.3%, Medium drops 24.8%, Ultra drops 38.5%, and even Low drops 15.9%. This steepening at higher resolutions is classic GPU-bound behavior: as pixel count increases by 2.25x from 1440p to 4K, the rendering load scales correspondingly, and the RTX 5080's 56.28 TFLOPS FP32 throughput, while enormous, cannot fully compensate for that pixel explosion when shader work is heavy. The data implies that 4K Ultra is the point where this combination finally meets its match in Thief, whereas 1080p and 1440p are well within the system's comfortable range.

GPU Role

The NVIDIA GeForce RTX 5080 brings 16 GB of GDDR7 memory on a 256-bit bus with 960.0 GB/s of bandwidth, and its boost clock sits at 2617 MHz with a base of 2295 MHz. In Thief, a 2014 stealth title, the GPU is clearly the dominant performance driver at 4K, where even the Low preset yields 127 FPS—still well above typical display refresh rates but far below the 172 FPS seen at 1080p. The memory subsystem is likely not a limiting factor here; Thief's textures and geometry from 2014 are unlikely to stress 16 GB of VRAM, and the 960.0 GB/s bandwidth is overkill for a game of that era. Instead, the bottleneck at 4K is raw shading and rasterization throughput.

The GPU's benchmark profile reinforces its position. It scores 36,565 in Passmark G3D, 255,450 in Geekbench Vulkan, and 235,901 in Geekbench OpenCL, with a Passmark GPU compute score of 21,789. Its percentile ranking among all GPUs is 87, meaning it outperforms 87% of tested graphics cards. Compared to its nearest rivals, the RTX 5080 sits just 0.6% ahead of the AMD Radeon 8060S and 0.7% ahead of the AMD Radeon RX 6750 GRE 12 GB, while trailing the AMD Radeon RX 9070 GRE by 2.2% in average benchmark score. These narrow deltas suggest that in synthetic tests, the RTX 5080 is competitive but not dominant; however, in Thief specifically, the measured FPS at 4K Ultra (75 FPS) and 4K High (79 FPS) show that the card's real-world gaming performance is strong, though the gap between High and Ultra at 4K is only 4 FPS, indicating diminishing returns from the highest preset.

The fact that 4K Ultra (75 FPS) barely differs from 4K High (79 FPS) while 4K Medium jumps to 103 FPS is telling. It implies that the Ultra preset in Thief includes effects that are disproportionately expensive relative to their visual payoff, and the GPU is spending significant resources on them. The 24 FPS difference between Medium and High at 4K (103 vs 79) is much larger than the 4 FPS difference between High and Ultra (79 vs 75), which suggests the High preset is already near the edge of the GPU's comfortable envelope at this resolution.

Settings Recommendations

The measured rows provide a clear hierarchy for optimizing the experience in Thief. At 1920x1080, the Low preset delivers 172 FPS, Medium delivers 155 FPS, High delivers 141 FPS, and Ultra delivers 138 FPS. The gap between High and Ultra is just 3 FPS, making Ultra the obvious choice at this resolution—the performance cost is negligible, and the visual quality is maximal. The Medium preset's 155 FPS comes with minimums of 132 FPS, which is still well above smoothness thresholds, but High's 141 FPS is a better balance if visual fidelity matters.

At 2560x1440, the story shifts slightly. Low runs at 151 FPS, Medium at 137 FPS (with minimums of 116 and maximums of 157), High at 124 FPS, and Ultra at 122 FPS. Again, High and Ultra are within 2 FPS of each other, so Ultra is the recommended pick for 1440p users—the 2 FPS cost is imperceptible. Medium's 137 FPS is 13 FPS higher than High, but the visual quality reduction may not justify that gain for most players. The minimum of 116 FPS on Medium ensures no stutter, but High's lack of a reported minimum makes it harder to assess its worst-case behavior.

At 3840x2160, the recommendations shift more decisively. Low runs at 127 FPS, Medium at 103 FPS (with minimums of 88 and maximums of 119), High at 79 FPS, and Ultra at 75 FPS. Here, Medium is the sweet spot: 103 FPS is substantially smoother than High's 79 FPS, and the 24 FPS gain is worth the visual downgrade from High to Medium. Ultra's 75 FPS is only 4 FPS below High, so if a player insists on maximum quality at 4K, Ultra is the better choice than High—the cost is minimal. However, the 88 FPS minimum on Medium ensures that even in demanding scenes, the frame rate stays above 60, which is critical for a stealth game where timing matters.

Across all resolutions, the data suggests that Ultra is never a poor choice at 1080p or 1440p, but at 4K, Medium offers the best balance of smoothness and visual quality. The 88 FPS minimum on 4K Medium is the key data point—it guarantees a playable experience, while High and Ultra at 4K lack reported minimums, introducing uncertainty about worst-case performance.

FAQ

Q: What is the highest average FPS this combination achieves in Thief?

A: The highest measured average is 172 FPS at 1920x1080 with Low settings. The next best is 155 FPS at 1080p Medium, followed by 151 FPS at 1440p Low.

Q: Is the system CPU-bound or GPU-bound at 4K?

A: The data indicates GPU-bound behavior at 4K. The frame rate drops significantly from 1440p to 4K across all settings, with Ultra falling from 122 FPS to 75 FPS (38.5% reduction), which is characteristic of the GPU struggling with pixel throughput rather than the CPU being unable to feed frames.

Q: How does the RTX 5080 compare to its nearest rival in synthetic benchmarks?

A: The RTX 5080's average benchmark score is 56,083, which is 0.6% higher than the AMD Radeon 8060S (55,757) and 0.7% higher than the AMD Radeon RX 6750 GRE 12 GB (55,698), but 2.2% lower than the AMD Radeon RX 9070 GRE (57,367).

Q: What is the FPS difference between High and Ultra settings at 1080p?

A: At 1920x1080, High averages 141 FPS while Ultra averages 138 FPS, a difference of just 3 FPS. This makes Ultra essentially free at this resolution.

Q: What is the minimum FPS recorded at 4K Medium?

A: The minimum FPS at 3840x2160 with Medium settings is 88 FPS, with a maximum of 119 FPS and an average of 103 FPS. This minimum is comfortably above 60 FPS, indicating smooth gameplay.

Q: Where does this combo rank among all tested combinations in Thief?

A: This combination ranks 347th out of 2,851 tested combos in Thief, placing it in the top 12.2% of all tested system configurations for this game.

How This Combo Ranks

The combination of the AMD Ryzen 5 7500X3D and NVIDIA GeForce RTX 5080 ranks 347th out of 2,851 tested combos in Thief, which places it in the top 12.2% of all systems tested for this game. This is a strong showing for a six-core CPU paired with a high-end GPU, but it is not at the absolute top of the charts. The ranking suggests that while the RTX 5080 provides exceptional GPU performance, the Ryzen 5 7500X3D's six-core, twelve-thread configuration may hold back the system in certain CPU-sensitive scenarios, particularly at lower resolutions where the GPU is not the bottleneck.

The CPU's average benchmark score is 44,573, which puts it at the 89th percentile among all CPUs. Its nearest rivals are the Intel Core Ultra X9 388H (44,466, 0.2% behind), the Intel Core i9-13950HX (44,342, 0.5% behind), the AMD Ryzen AI Max 385 (44,309, 0.6% behind), and the Intel Core i5-14600 (44,889, 0.7% ahead). These are all within 0.7% of each other, indicating that the Ryzen 5 7500X3D is competitive with a broad range of high-end mobile and desktop processors in synthetic benchmarks. Its Passmark single-thread score of 3,494 and multithread score of 25,047 suggest balanced performance, though the six cores limit its multithreaded ceiling compared to higher-core-count rivals.

The 347th rank out of 2,851 combos means that about 88% of tested systems perform worse in Thief, but it also implies that a substantial number of combos outperform this one, likely with higher-core-count CPUs that can push frame rates higher at 1080p. The measured data supports this: at 1080p Low, the system hits 172 FPS, but that is likely CPU-limited, as the RTX 5080 could theoretically render far more frames if the CPU could feed it faster.

CPU Role

The AMD Ryzen 5 7500X3D is a six-core, twelve-thread processor based on the Raphael codename and Zen 4 architecture, built on a 5 nm process at TSMC. Its base clock is 4.00 GHz with a boost clock of 4.50 GHz, and it features 96 MB of shared L3 cache—a substantial pool that benefits gaming workloads. The CPU's TDP is 65 watts, and it supports DDR5 memory with a dual-channel bus and 83.2 GB/s of bandwidth. Its Passmark multithread score is 25,047, and its single-thread score is 3,494.

In Thief, the CPU's role becomes apparent when examining the FPS scaling between 1080p and 1440p. At Low settings, the frame rate drops from 172 FPS at 1080p to 151 FPS at 1440p—a 12.2% reduction. If the system were purely GPU-bound at 1080p, this drop would be larger, as the GPU would be the limiting factor. The relatively modest drop suggests that at 1080p Low, the CPU is partially limiting performance, preventing the GPU from reaching even higher frame rates. The Ryzen 5 7500X3D's six cores and 4.50 GHz boost clock are strong for gaming, but they cannot match the frame-feeding capability of higher-core-count processors in lightly-threaded scenarios where Thief's engine may not scale perfectly across twelve threads.

The CPU's benchmark data supports its capability. Its floating-point math score is 43,594, integer math is 70,774, and extended instructions score is 20,455. These numbers indicate solid computational throughput for a six-core part, but they also reveal that the CPU is not a top-tier multithreaded performer—its Passmark multithread score of 25,047 is dwarfed by high-core-count desktop processors. In Thief, a 2014 title that likely does not utilize more than a few threads efficiently, the single-thread performance (3,494) is more critical, and here the CPU performs admirably, as evidenced by the healthy frame rates at 1080p.

The 96 MB L3 cache is a notable advantage for gaming, as it reduces memory latency and improves hit rates for game data. This likely contributes to the CPU's ability to maintain high frame rates in Thief, even at 1080p where the GPU is less stressed. The 83.2 GB/s memory bandwidth is adequate for a dual-channel DDR5 setup, though it is not exceptional compared to higher-end platforms.

Measured FPS Breakdown

At 1920x1080, the measured average FPS values are: Low at 172 FPS, Medium at 155 FPS with a minimum of 132 FPS and maximum of 179 FPS, High at 141 FPS, and Ultra at 138 FPS. The spread from Low to Ultra is 34 FPS, and the gap between High and Ultra is just 3 FPS. Medium's minimum of 132 FPS ensures no dips below smoothness thresholds, while Low's 172 FPS is the highest recorded average in the entire dataset.

At 2560x1440, the averages are: Low at 151 FPS, Medium at 137 FPS with a minimum of 116 FPS and maximum of 157 FPS, High at 124 FPS, and Ultra at 122 FPS. The spread from Low to Ultra is 29 FPS, slightly narrower than at 1080p. The gap between High and Ultra is 2 FPS, making them nearly indistinguishable in performance. Medium's minimum of 116 FPS is still well above 60 FPS, indicating smooth gameplay.

At 3840x2160, the averages are: Low at 127 FPS, Medium at 103 FPS with a minimum of 88 FPS and maximum of 119 FPS, High at 79 FPS, and Ultra at 75 FPS. The spread from Low to Ultra is 52 FPS, the widest of any resolution, reflecting the GPU's increasing workload. The gap between High and Ultra is 4 FPS, while the gap between Medium and High is 24 FPS—a significant jump that makes Medium the most attractive option at 4K. Medium's minimum of 88 FPS is the lowest recorded minimum in the dataset, but it remains comfortably above 60 FPS, ensuring a playable experience.

The data also reveals that the maximum FPS values are only reported for Medium settings at each resolution: 179 FPS at 1080p, 157 FPS at 1440p, and 119 FPS at 4K. These maximums indicate that the system can briefly reach much higher frame rates than the averages suggest, particularly at lower resolutions. The absence of min/max data for other settings means that only Medium gives a complete picture of frame time variance, and its minimums (132, 116, and 88 FPS respectively) show consistent performance without severe dips.