Thief

The AMD Ryzen 7 7800X3D and NVIDIA GeForce RTX 5050 combination delivers a clear performance picture in Thief (2014), a stealth title whose engine responds predictably to both CPU and GPU resources. Across the measured resolutions, the pairing shows a distinct shift from CPU-limited high frame rates at 1080p to GPU-bound constraints at 4K, with the 8-core/16-thread processor providing a stable foundation while the 8 GB graphics card defines the ceiling at higher fidelity settings.

CPU Role — cores, clocks, and how they relate to this game's results

The Ryzen 7 7800X3D brings 8 cores and 16 threads to the test bench, built on the Zen 4 (Raphael) architecture with a base clock of 4.20 GHz and a boost clock of 5.00 GHz. This processor posts a single-thread score of 959 in 3DMark and 2,426 in Geekbench, alongside a multi-thread 3DMark score of 7,980 and a Cinebench R23 multi-core result of 16,817. For Thief, a title from 2014 that relies on moderate thread scaling, the 7800X3D’s combination of high single-core throughput and substantial L3 cache (96 MB shared) ensures that frame pacing remains consistent even when the GPU is not the bottleneck.

At 1080p, the CPU’s influence is most visible. The measured 164 FPS average at Low settings and 124 FPS at Medium settings indicate that the processor can feed the RTX 5050 well above the 90–95 FPS range seen at High and Ultra, respectively. The delta between Low and Ultra at 1080p is 74 FPS (164 vs. 90), which is a 45% drop driven by graphical load rather than CPU limitation. In contrast, the 3DMark 2-thread score of 1,887 and 4-thread score of 3,684 suggest that Thief's engine—which typically uses 4–8 threads—finds ample headroom in the 7800X3D’s 16 threads. The processor’s 83rd percentile ranking among all CPUs, with an average benchmark score of 33,079, places it just 0.4% behind the AMD Ryzen 9 PRO 6950H (33,201) and statistically tied with the Ryzen 7 8700G and Core i7-13650HX (both at 33,089). This near-parity in synthetic scores translates to negligible CPU-side differences in Thief’s frame delivery; the data shows that the processor is never the primary limiter at any resolution tested.

FAQ — Q&A pairs answerable from FACT PACK data

*Q: What is the highest average FPS this combo achieves in Thief, and at which settings?*

A: The highest measured average is 164 FPS at 1920x1080 with Low settings. The next closest is 124 FPS at the same resolution with Medium settings, followed by 112 FPS at 2560x1440 on Low.

Q: How does the CPU compare to its nearest rivals in synthetic benchmarks?

A: The Ryzen 7 7800X3D’s average benchmark score of 33,079 is within 0.4% of the AMD Ryzen 9 PRO 6950H (33,201) and effectively identical to the AMD Ryzen 7 8700G and Intel Core i7-13650HX (both 33,089). This means the CPU’s performance tier is consistent with those alternatives in general-purpose compute.

*Q: Is the RTX 5050’s 8 GB VRAM a limiting factor in Thief?*

A: The data does not show VRAM capacity as a direct bottleneck. Measured FPS at 4K Ultra averages 36 FPS, which is more consistent with the GPU’s raw compute limits (13.17 TFLOPS FP32) than with memory exhaustion. The GPU’s 320.0 GB/s bandwidth and 128-bit bus are the relevant constraints.

Q: What frame rate can users expect at 1440p with High settings?

A: At 2560x1440 with High settings, the combo averages 65 FPS. This is 31.6% lower than the 1080p High result of 95 FPS, indicating a clear resolution scaling penalty.

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

A: The combo ranks 1,710 out of 2,851 tested combinations, placing it in the upper 40% of all configurations. This rank reflects the balance between the CPU’s high percentile (83rd) and the GPU’s mid-tier percentile (66th).

Q: What is the largest FPS gap between settings at a single resolution?

A: At 3840x2160, the gap between Low (66 FPS) and Ultra (36 FPS) is 30 FPS, a 45.5% reduction. At 1080p, the same settings gap is 74 FPS (164 vs. 90), which is a larger absolute drop but a similar relative percentage.

Resolution Scaling — how FPS drops from 1080p to 4K and what it says about the limiting component

The measured FPS progression across resolutions reveals a textbook GPU-bound scaling pattern. At 1080p High, the combo delivers 95 FPS; moving to 1440p High drops that to 65 FPS (a 31.6% reduction), and 4K High falls to 38 FPS (a 60% reduction from 1080p). This steep decline is consistent with a GPU that is approaching its limits as pixel count rises. The RTX 5050’s 13.17 TFLOPS FP32 throughput and 82.30 GPixel/s pixel rate are the primary drivers of this behavior—each resolution step increases the pixel workload by roughly 1.78x (1080p to 1440p) and 2.25x (1440p to 4K), and the FPS drops track accordingly.

The Low settings row provides the clearest view of the CPU’s headroom. At 1080p Low, the combo hits 164 FPS, while 1440p Low yields 112 FPS and 4K Low yields 66 FPS. The 1080p-to-1440p drop of 52 FPS (31.7%) and the 1440p-to-4K drop of 46 FPS (41.1%) show that even at minimal graphical load, the GPU becomes the limiter at higher resolutions. If the CPU were the bottleneck, the FPS would remain flat or drop only slightly as resolution increases—but the data shows a consistent decline, confirming that the 7800X3D’s 16 threads and 5.00 GHz boost clock are sufficient for Thief’s engine at all tested settings.

The Medium settings row includes min/max FPS values, offering insight into frame consistency. At 1080p Medium, the minimum is 106 FPS and the maximum is 143 FPS, a range of 37 FPS around the 124 FPS average. At 1440p Medium, the range narrows to 26 FPS (72–98 FPS around an 85 FPS average), and at 4K Medium, the range is 15 FPS (43–58 FPS around a 50 FPS average). This tightening of the frame time distribution as resolution increases further supports the GPU-bound interpretation—higher pixel counts smooth out CPU-related spikes because the GPU’s rendering time dominates the frame budget.

How This Combo Ranks — use comboRankInGame vs other tested combos

The combo’s rank of 1,710 out of 2,851 tested combinations places it in the 60th percentile of all configurations tested in Thief. This is a mid-pack result that aligns with the component profiles: the CPU’s 83rd percentile standing is offset by the GPU’s 66th percentile, yielding a combined position that is slightly below the midpoint of the distribution. The rank suggests that while the 7800X3D is a top-tier processor for CPU-bound scenarios, the RTX 5050’s mid-range compute capabilities (13.17 TFLOPS) prevent this pairing from reaching the upper echelons of the ranking table.

Comparing to the GPU’s nearest rivals provides context. The RTX 5050’s average benchmark score of 21,035 is within 1.6% of the AMD Radeon RX 5600 XT (20,713), and within 1% of the AMD Radeon HD 8970M (21,237). The RTX 5050 sits 0.6% below the AMD Radeon RX Vega M GL (21,153) and 1.6% below the NVIDIA RTX A4000 Mobile (21,379). These margins are small enough that in Thief, the differences would manifest as a few FPS at most. The 1710th rank reflects this tight clustering—the combo is not an outlier but rather sits within a dense band of similar-performing graphics solutions, with the CPU providing a marginal edge over weaker processors in the same GPU tier.

GPU Role — VRAM, clocks, and how they relate to this game's results

The RTX 5050 is built on the Blackwell 2.0 architecture with a GB207 chip, featuring 2,560 shading units, 80 TMUs, and 32 ROPs. Its base clock of 2317 MHz boosts to 2572 MHz, with memory running at 2500 MHz (20 Gbps effective) across an 8 GB GDDR6 interface with a 128-bit bus. The resulting 320.0 GB/s memory bandwidth and 13.17 TFLOPS FP32 performance define the GPU’s capability in Thief. The 82.30 GPixel/s pixel rate and 205.8 GTexel/s texture rate are directly relevant to a 2014 title that relies on traditional rasterization rather than ray tracing or tensor core workloads.

The 8 GB VRAM capacity is sufficient for Thief’s textures and geometry at all tested settings, as evidenced by the fact that 4K Ultra still produces a playable 36 FPS average rather than stalling due to memory overflow. The GPU’s 66th percentile ranking (average benchmark score 21,035) places it in the mid-range of all graphics cards, and its performance in Thief reflects this—at 1080p, it delivers 90–164 FPS depending on settings, which is a strong result for a 2014 game. However, the 4K results (36–66 FPS across settings) show that the GPU’s compute throughput becomes the hard ceiling at higher resolutions, particularly at Ultra where the 36 FPS average is below the 60 FPS threshold many users consider smooth.

The GPU’s direct competitors in synthetic benchmarks—the RX 5600 XT, RX Vega M GL, HD 8970M, and RTX A4000 Mobile—all fall within a 1.6% band of the RTX 5050’s average score. In Thief, this translates to sub-5 FPS differences at any given resolution and settings combination, assuming identical CPUs. The RTX 5050’s 20 RT cores and 80 tensor cores are present but unused in this title, meaning the 13.17 TFLOPS FP32 figure is the operative specification for frame generation.

Measured FPS Breakdown — resolution by resolution, settings by settings, exact numbers

At 1920x1080, the combo produces its highest frame rates. Low settings yield 164 FPS average, Medium 124 FPS (with a min of 106 and max of 143), High 95 FPS, and Ultra 90 FPS. The gap between Low and Ultra is 74 FPS, and the step from High to Ultra is a modest 5 FPS (95 to 90), suggesting that Ultra’s additional effects (likely ambient occlusion or shadows) impose a minor cost at this resolution.

At 2560x1440, the averages drop to 112 FPS on Low, 85 FPS on Medium (min 72, max 98), 65 FPS on High, and 62 FPS on Ultra. The High-to-Ultra delta narrows to 3 FPS, indicating that the GPU’s fill rate limits begin to dominate. The Low-to-Ultra gap is 50 FPS (112 vs. 62), which is a 44.6% reduction—similar to the 1080p percentage drop.

At 3840x2160, the averages are 66 FPS on Low, 50 FPS on Medium (min 43, max 58), 38 FPS on High, and 36 FPS on Ultra. The Low-to-Medium drop is 16 FPS (24.2%), Medium-to-High is 12 FPS (24%), and High-to-Ultra is only 2 FPS (5.3%). This final small delta reveals that at 4K, the GPU is so heavily pixel-bound that the difference between High and Ultra presets becomes negligible—the RTX 5050 simply cannot render the additional detail fast enough to create a meaningful gap.

The Medium settings row at each resolution is the only one with complete min/max data. At 1080p, the spread is 37 FPS (106–143); at 1440p, it narrows to 26 FPS (72–98); and at 4K, it narrows further to 15 FPS (43–58). This pattern indicates that as resolution increases, frame time variance decreases relative to the average, which is characteristic of a GPU-bound scenario where rendering time per frame is long enough to absorb transient CPU delays.

Settings Recommendations — which preset gives the best experience per the measured rows

For users targeting 60 FPS as a playable baseline, the data supports 1440p High as the optimal choice. The 65 FPS average at this setting exceeds the 60 FPS target, and the 1440p resolution provides a meaningful visual upgrade over 1080p. At 1080p, all settings exceed 90 FPS, so users with 144 Hz or higher refresh rate monitors should consider High (95 FPS) or Ultra (90 FPS) to maximize visual fidelity while staying above 90 FPS. The Low preset at 1080p (164 FPS) is only useful for competitive play or extreme refresh rate displays.

At 4K, the situation is more constrained. The only setting that clears 60 FPS is Low (66 FPS), while Medium (50 FPS) and High (38 FPS) fall short. For 4K users, Low is the only preset that delivers a consistently smooth experience, and even then, the 66 FPS average carries the risk of dips below 60 FPS during intense scenes—the Medium min of 43 FPS at 4K hints at the frame time variance that could affect Low as well, though no min is provided for Low. Ultra at 4K (36 FPS) is only suitable for screenshot-style play or cinematic viewing, not interactive stealth gameplay.

The best overall experience per the measured rows is 1440p High. It offers a 65 FPS average, which is 72% of the 1080p High score but at a resolution with 1.78x more pixels. The step up to 1440p Ultra (62 FPS) gains only 3 FPS over High, making High the better choice—the Ultra preset’s additional graphical effects do not justify the marginal frame rate cost. For users who prioritize frame rate over resolution, 1080p Ultra (90 FPS) is the highest-fidelity option that stays above 90 FPS, while 1080p Medium (124 FPS) provides a 37% frame rate boost over Ultra with only a single preset drop in quality.