Thief

The AMD Ryzen 7 7800X3D paired with the NVIDIA GeForce RTX 2060 produces a highly instructive dataset for Thief, a 2014 stealth title. The benchmark results reveal a system where the GPU is the overwhelming performance governor, with the CPU’s considerable headroom remaining largely untapped across all tested configurations. The data indicates a clear hierarchy: as resolution increases, the frame rate collapses in a pattern typical of a graphics-card-bound scenario.

Resolution Scaling

The most striking trend in the measured FPS is the dramatic performance drop as resolution scales from 1920x1080 to 3840x2160. At High settings, the average frame rate at 1080p is 79.4 FPS. Moving to 2560x1440, this falls to 55.1 FPS, a reduction of roughly 30%. At 4K, the average plummets to 32.9 FPS, which is less than half of the 1080p figure. This steep decline is a classic indicator that the rendering pipeline is saturated by the GPU’s pixel throughput, not by the CPU’s ability to feed draw calls.

The pattern holds consistently across all quality presets. For example, at Medium settings, the system delivers 107.4 FPS at 1080p, 73.7 FPS at 1440p, and 44.6 FPS at 4K. The scaling factor from 1080p to 4K is nearly identical across Low, Medium, and High settings, suggesting a uniform bottleneck. The only outlier is Ultra, where 1080p yields 78 FPS, 1440p yields 55.8 FPS, and 4K yields 30.2 FPS. The 4K Ultra result is the lowest of any configuration, confirming that the RTX 2060’s compute and memory subsystems are the limiting factors.

The Ryzen 7 7800X3D’s capabilities, as evidenced by its 3DMark scores (7980 in 16-thread, 959 in single-thread), are far beyond what Thief demands. The game’s engine, released in 2014, cannot leverage the CPU’s 8 cores and 16 threads effectively. Consequently, resolution scaling has no inverse effect on CPU load that would mitigate the FPS drop; instead, the GPU’s fixed workload per pixel dictates the outcome. The data shows that 1080p is the only resolution where the RTX 2060 can consistently exceed 100 FPS, and even then only at Low and Medium settings.

GPU Role

The NVIDIA GeForce RTX 2060, with its 6 GB of GDDR6 memory on a 192-bit bus, provides a bandwidth of 336.0 GB/s. This memory configuration is a pivotal factor in the 4K results. At 3840x2160, the frame buffer demand for High and Ultra settings likely approaches the card’s capacity limits. The average FPS at 4K Ultra is 30.2, and at 4K High it is 32.9, indicating that the card is struggling to maintain playable rates. The clock speeds, base 1365 MHz and boost 1680 MHz, are modest by modern standards, and the data suggests they are insufficient for high-resolution rasterization in this title.

The GPU’s FP32 performance of 6.451 TFLOPS is the primary driver of the observed results. At 1080p Low, the system hits 139.4 FPS, which is over four times the 4K Ultra figure. This disparity shows that the RTX 2060 has ample raw throughput for low-resolution work but lacks the fill rate and memory bandwidth for 4K. The pixel rate of 80.64 GPixel/s and texture rate of 201.6 GTexel/s are also relevant; Thief’s shadow and lighting effects at Ultra settings place a heavy burden on these units.

Notably, the GPU’s percentile rank against all GPUs is 58, placing it in the lower-mid tier. Its nearest rivals, such as the AMD Radeon RX 7600S with a 0.5% higher average score, confirm that this is not a high-end part. In Thief, the data shows that the RTX 2060 is the definitive bottleneck at every resolution above 1080p. The CPU’s 96 MB of shared L3 cache and high single-thread performance (Cinebench R23 single-core score of 4115) cannot compensate for the GPU’s limitations. The measured FPS values are entirely consistent with a GPU that is maxed out, not a CPU that is waiting on data.

FAQ

Q: Why does the frame rate at 4K High (32.9 FPS) differ so little from 4K Ultra (30.2 FPS)?

A: The data shows only a 2.7 FPS difference between these two settings. This narrow gap indicates that the RTX 2060’s memory bandwidth and pixel fill rate are already saturated at High, so the additional rendering work in Ultra cannot be processed efficiently. The GPU is the limiting component, not the settings.

Q: Is the AMD Ryzen 7 7800X3D holding back the RTX 2060 in this game?

A: No. Benchmark results indicate the CPU is far more powerful than required. The CPU’s 3DMark 16-thread score of 7980 and Passmark single-thread score of 3759 are far above what a 2014 game like Thief can utilize. The FPS scaling from 1080p to 4K is purely a function of GPU load, proving the CPU is not a bottleneck.

Q: What is the best resolution for a smooth experience with this combo?

A: Based on the measured data, 1920x1080 is the only resolution where the average FPS remains above 60 across all settings. At 1080p Medium, the system achieves 107.4 FPS, and at 1080p High it achieves 79.4 FPS. Dropping to 1440p reduces the average below 100 FPS even at Low settings (96.7 FPS).

Q: How does the RTX 2060 compare to its closest rivals in synthetic tests?

A: The GPU’s average benchmark score is 16074, placing it between the AMD Radeon Pro 5600M (16351, which is 1.7% higher) and the AMD Radeon RX 7600S (15996, which is 0.5% lower). These small deltas indicate the RTX 2060 performs in a tight competitive band, though its real-world Thief results are what matter here.

Q: Does the 6 GB VRAM capacity cause any specific issues at 4K?

A: The data suggests it may. At 4K Ultra, the average FPS is 30.2, and at 4K High it is 32.9. The minimal difference between these two presets implies that the card is either hitting a VRAM capacity wall or a bandwidth limit, preventing it from leveraging the higher-quality textures. The 336.0 GB/s bandwidth is likely the primary constraint.

Q: Why is the 1080p Low result (139.4 FPS) so much higher than the 1440p Low result (96.7 FPS)?

A: This 42.7 FPS drop is due to the exponential increase in pixel count. The RTX 2060’s fill rate of 80.64 GPixel/s is sufficient for 1080p but becomes a limiting factor at 1440p. The CPU’s performance is irrelevant here, as the GPU’s pixel processing capability dictates the output.

Measured FPS Breakdown

The complete dataset for the AMD Ryzen 7 7800X3D and NVIDIA GeForce RTX 2060 combination is as follows. At 1920x1080, the system delivers an average of 139.4 FPS on Low settings. Medium settings reduce this to 107.4 FPS. High settings yield 79.4 FPS, and Ultra settings produce 78 FPS. The difference between High and Ultra at 1080p is minimal (1.4 FPS), suggesting a CPU or driver-level cap rather than a GPU limitation at those settings.

At 2560x1440, the averages drop significantly. Low settings achieve 96.7 FPS, which is still playable but 30.6% lower than the 1080p Low result. Medium settings yield 73.7 FPS, High settings 55.1 FPS, and Ultra settings 55.8 FPS. Notably, the Ultra result at 1440p is slightly higher than High, which is an anomaly in the data, likely due to measurement variance or specific scene complexity, but the overall trend is downward.

At 3840x2160, the card struggles. Low settings produce 58.6 FPS, which is the only 4K result above 60 FPS. Medium settings drop to 44.6 FPS, High settings to 32.9 FPS, and Ultra settings to 30.2 FPS. The scaling from 1440p to 4K at High settings is a 40% reduction in average FPS. The data clearly shows that 4K is only viable with Low settings, and even then, the margin above 60 FPS is slim.

How This Combo Ranks

The combination of the Ryzen 7 7800X3D and RTX 2060 ranks 1374 out of 1642 tested combos in Thief. This places it in the lower 16th percentile of all configurations, which is a poor showing for a system with such a capable processor. The rank is heavily influenced by the GPU’s mid-tier status; the RTX 2060’s percentile rank of 58 against all GPUs is dragged down by the fact that Thief is not CPU-bound.

The CPU’s performance relative to its rivals is excellent. The Ryzen 7 7800X3D has an average benchmark score of 31580, which is 0.1% higher than the Intel Core i7-12700F (31543) and 0.3% higher than the AMD Ryzen 9 5980HX (31495). However, this CPU strength is irrelevant to the combo’s ranking because the GPU cannot translate it into higher frame rates. The data shows that the combo’s rank is determined by the GPU’s absolute performance in the game, not by the CPU’s synthetic capabilities.

In practical terms, this rank means that a significant majority of other CPU-GPU combinations tested will outperform this setup in Thief. The 1374th position out of 1642 indicates that while the system is not the worst, it is far from optimal for this title. The bottleneck is unambiguous: the RTX 2060’s 6 GB memory and 6.451 TFLOPS FP32 throughput are insufficient for the higher resolutions, while the CPU sits idle. For users targeting 1080p, the experience is adequate, but for higher resolutions, the data suggests that a GPU upgrade would be the only meaningful change to improve the ranking.