Thief

# Thief — AMD Ryzen 5 5600 + NVIDIA GeForce RTX 2070 SUPER

The measured FPS data for this combination reveals a system that scales predictably across resolutions and settings, with the RTX 2070 SUPER acting as the primary performance gate at higher presets while the Ryzen 5 5600 comfortably feeds frames at lower settings. This pairing, ranked 988th out of 1642 tested combos in Thief, shows a 1080p-to-4K transition that highlights GPU-bound behavior in a 2014 stealth title with modern hardware.

Resolution Scaling

The FPS drop from 1920x1080 to 3840x2160 is substantial but not uniform across settings, indicating where the bottleneck shifts. At Low settings, the average FPS falls from 160.5 at 1080p to 117.7 at 1440p (a 26.7% reduction) and then to 68.8 at 4K (a 57.1% drop from 1080p). This steep 4K penalty at Low suggests the GPU’s pixel throughput becomes the limiting factor even when shader work is minimal. The RTX 2070 SUPER’s 64 ROPs and 113.3 GPixel/s pixel rate are the likely constraints here, as fill-rate-bound tasks dominate at 4K.

At High settings, the scaling is more gradual: 100.5 FPS at 1080p, 68.8 at 1440p (31.6% drop), and 39.3 at 4K (60.9% drop). The near-identical 4K High and 1440p Low scores (39.3 vs. 68.8) underscore that resolution scaling at 4K roughly halves performance regardless of preset, pointing to a consistent GPU limit rather than a CPU stall. The Ultra preset shows the largest absolute loss: 92 FPS at 1080p falls to 37.3 at 4K, a 59.5% reduction. This pattern—where 1080p-to-1440p losses range from 26.7% to 31.3% and 1440p-to-4K losses range from 40.9% (Medium) to 46.3% (Low)—indicates that the 4K jump is disproportionately expensive, typical of a GPU with 8 GB GDDR6 and 448.0 GB/s bandwidth handling pixel-heavy loads.

GPU Role

The RTX 2070 SUPER’s 8 GB GDDR6 memory on a 256-bit bus with 448.0 GB/s bandwidth is sufficient for Thief’s 2014-era assets, yet the data shows VRAM is not the constraining factor—fill rate and shader throughput are. The GPU’s base clock of 1605 MHz and boost clock of 1770 MHz, paired with 2560 shading units and 160 TMUs, produce a 9.062 TFLOPS FP32 rate. At 4K High, the 39.3 FPS average suggests the GPU is nearing its compute ceiling; the 283.2 GTexel/s texture rate is likely saturated by Thief’s detailed environments at that resolution.

The passmark G3D score of 18169 places this GPU in the 65th percentile of all GPUs, with nearest rivals including the NVIDIA RTX A400 (avgScore 22307, deltaPct -0.5) and AMD Radeon RX 5700 (avgScore 21670, deltaPct 2.5). The data reveals that at 4K Ultra, the 37.3 FPS average is only 5.1% below the 4K High result, implying diminishing returns from the Ultra preset’s added effects—likely shadow and post-processing loads—rather than a VRAM spillover. The GPU’s 12 nm Turing process and 215 W TDP suggest thermal headroom is adequate for sustained loads, but the measured FPS curve shows no evidence of clock throttling, as 1080p Low achieves 160.5 FPS, which would expose any CPU-side stutter.

CPU Role

The Ryzen 5 5600’s 6 cores and 12 threads from the Zen 3 (Vermeer) architecture, clocked at 3.50 GHz base and 4.40 GHz boost, provide ample headroom for Thief’s single-threaded-leaning engine. The CPU’s 3dmark single-thread score of 882 and multi-thread score of 5641 (16 threads) indicate strong per-core performance, which matters for a 2014 stealth game that rarely scales beyond a few threads. The 32 MB shared L3 cache and 51.2 GB/s memory bandwidth are more than sufficient for this workload, as evidenced by the 1080p Low result of 160.5 FPS—a figure that would be CPU-bound if the processor were weak, but here it reflects a balanced system.

The CPU’s nearest rivals in synthetic benchmarks—Intel Xeon D-1746TER (avgScore 21635, deltaPct 0.6) and Intel Core i7-11700F (avgScore 21957, deltaPct -0.9)—show the Ryzen 5 5600 sits within 1% of these parts, meaning any performance differences in Thief are attributable to GPU scaling, not CPU capability. The 79th percentile ranking across all CPUs reinforces that this processor is not the bottleneck in any measured scenario. At 1440p Low, the 117.7 FPS average is 36.5% lower than 1080p Low, which aligns with GPU pixel-rate limits rather than CPU thread saturation. The data implies that even at 1080p High (100.5 FPS), the CPU is delivering frames faster than the GPU can process, as the 4K High score drops to 39.3 FPS without any sign of CPU-induced frame pacing issues.

Settings Recommendations

The measured rows suggest that the Medium preset at 1440p offers the best balance for this hardware. At 2560x1440 Medium, the 86.5 FPS average is 25.7% higher than High (68.8 FPS) while being only 9.1% below the 1080p Medium result (130.9 FPS), indicating that most visual features scale efficiently. For 4K users, Low at 68.8 FPS is the only playable option, as Medium drops to 54.3 FPS and High falls to 39.3 FPS—both below the 60 FPS threshold that the Low preset maintains. At 1080p, Ultra’s 92 FPS is 8.5% lower than High’s 100.5 FPS, making High the pragmatic choice unless maximum fidelity is prioritized, since the 8.5 FPS gain from dropping to Medium (130.9 FPS) yields diminishing visual returns.

The Ultra preset across all resolutions shows the smallest margins: 92 FPS at 1080p, 63.2 at 1440p, and 37.3 at 4K. These figures suggest Ultra is only worthwhile at 1080p, where the 92 FPS average remains competitive. For 1440p, Medium’s 86.5 FPS is 36.9% higher than Ultra’s 63.2 FPS, a substantial playability gain for likely minor shadow and texture quality differences. The data strongly implies that this GPU’s sweet spot is 1440p Medium, as it delivers 86.5 FPS—a comfortable margin above 60 FPS—while 4K requires dropping to Low, sacrificing too much visual quality for the resolution’s benefit.

FAQ

Q: Can this combo run Thief at 4K smoothly?

A: The data shows 4K is only playable at Low settings (68.8 FPS average). Medium (54.3 FPS) and High (39.3 FPS) fall below 60 FPS, and Ultra (37.3 FPS) is the lowest of all measured points.

Q: What is the best resolution for this system?

A: 2560x1440 at Medium settings yields 86.5 FPS, which is 25.7% higher than High at the same resolution and remains above 60 FPS. This is the most balanced configuration in the measured data.

Q: How does the CPU compare to its closest rivals?

A: The Ryzen 5 5600’s average benchmark score is 21765, which is 0.6% higher than the Intel Xeon D-1746TER and 0.9% higher than the Intel Core i7-11700F, placing it in the 79th percentile of all CPUs.

Q: Is the GPU or CPU the bottleneck at 1080p High?

A: The 1080p High average of 100.5 FPS versus 1080p Low’s 160.5 FPS indicates a 37.4% performance gap due to GPU load, not CPU limits—the processor’s single-thread score of 882 suggests it is not the limiting factor.

Q: What is the FPS difference between High and Ultra at 1440p?

A: High averages 68.8 FPS while Ultra averages 63.2 FPS, a difference of 5.6 FPS (8.1%). This small gap suggests Ultra’s extra effects have minimal impact at this resolution.

Q: How does this GPU rank against all tested GPUs?

A: The RTX 2070 SUPER scores 18169 on passmark G3D, placing it in the 65th percentile, with its nearest rival being the AMD Radeon RX 5700 (avgScore 21670, deltaPct 2.5).

Measured FPS Breakdown

At 1920x1080, the settings ladder is: Low 160.5 FPS, Medium 130.9 FPS, High 100.5 FPS, Ultra 92 FPS. The gap between Low and Ultra is 68.5 FPS (42.7%), showing significant headroom for visual upgrades. At 2560x1440, Low drops to 117.7 FPS, Medium to 86.5 FPS, High to 68.8 FPS, and Ultra to 63.2 FPS—a 54.5 FPS spread (46.3%). The 3840x2160 results are: Low 68.8 FPS, Medium 54.3 FPS, High 39.3 FPS, Ultra 37.3 FPS, with a 31.5 FPS spread (45.8%). Notably, the 4K Low score (68.8 FPS) exactly matches the 1440p High score (68.8 FPS), suggesting these two configurations impose identical GPU workloads. The 1080p Ultra score of 92 FPS is 34.9% higher than the 4K Low score, illustrating the resolution penalty even at the lowest settings. Across all presets, the average FPS drop from 1080p to 1440p is 30.7%, while the drop from 1440p to 4K is 44.4%, confirming that the 4K transition is the more severe bottleneck for this hardware.

How This Combo Ranks

This AMD Ryzen 5 5600 + NVIDIA GeForce RTX 2070 SUPER combination ranks 988th out of 1642 tested combos in Thief, placing it in the 39.8th percentile of all system pairings. The CPU’s 79th percentile ranking versus the GPU’s 65th percentile suggests the graphics card is the weaker link in this pairing, which the FPS data confirms through consistent GPU-bound behavior at higher resolutions. The combo’s mid-pack ranking reflects that while the CPU is competitive (within 1% of its nearest rivals), the GPU’s 8 GB VRAM and 9.062 TFLOPS compute are modest by modern standards, as evidenced by the 4K Ultra score of 37.3 FPS. Relative to other 1642 configurations, this system is positioned for 1080p and 1440p gaming, with the 1080p High score of 100.5 FPS representing a strong result, while the 4K Ultra score of 37.3 FPS shows clear limitations. The rank of 988 implies that over 600 tested combos outperform this pairing in Thief, likely due to newer GPUs with higher pixel rates and more shading units, but the data also shows this combo is far from the bottom—the 4K Low result of 68.8 FPS indicates playable performance exists even at the highest resolution with reduced settings.