Thief

The pairing of the AMD Ryzen 5 2600 with the NVIDIA GeForce RTX 5090 in Thief produces a dataset that is defined by a clear bottleneck: the processor. While the RTX 5090 is a modern, high-bandwidth graphics card, the measured frame rates across all resolutions suggest that the 2018-era six-core CPU is the limiting factor. This analysis walks through the resolution scaling, settings behavior, and component roles to explain why this combination performs the way it does.

Resolution Scaling

The measured FPS data shows an unusual pattern: performance does not scale dramatically with resolution, which indicates the CPU is the primary constraint. At 1080p with Low settings, the combo achieves 110.1 FPS. Moving to 1440p at Low settings only drops this to 95.2 FPS, a reduction of roughly 13.5%. At 4K with Low settings, the average is 83.8 FPS, which is still a relatively modest drop from the 1440p figure. This is the signature of a CPU-bound scenario: if the GPU were the limiting factor, the leap from 1080p to 4K would typically produce a much steeper decline in frame rate due to the massive increase in pixel count.

The gap between the lowest and highest resolution settings narrows further when looking at the High preset. At 1080p High, the average is 88 FPS; at 1440p High, it is 79 FPS; and at 4K High, it is 64.7 FPS. The delta between 1080p and 4K here is about 26.5%, which is still relatively small for a game of this era. The RTX 5090, with its 32 GB of GDDR7 memory and 1.79 TB/s bandwidth, is clearly not being saturated by this workload. Instead, the Ryzen 5 2600, with its 6 cores and 12 threads, appears to be the ceiling. The benchmark results indicate that even at 4K, the CPU cannot feed the GPU fast enough to push frame rates much higher than the 84 FPS observed at Low settings.

For context, the RTX 5090 sits in the 94th percentile of all GPUs with an average benchmark score of 84306, while the Ryzen 5 2600 is in the 77th percentile for CPUs with an average score of 19616. This pairing of a top-tier GPU with a mid-range older CPU creates a situation where resolution scaling is muted. The data suggests that the GPU is waiting on the CPU at every resolution, and the frame rate differences between 1080p and 4K are primarily due to the slight increase in GPU workload, not a fundamental shift in the bottleneck.

Settings Recommendations

The measured rows provide a clear picture of how settings affect performance, and the optimal choice depends on the target resolution. At 1080p, the Low preset yields 110.1 FPS, which is the highest measured value in the entire dataset. However, the jump from Low to Medium costs 12.4 FPS (97.7 FPS), and the jump from Medium to High costs another 9.7 FPS (88 FPS). The Ultra preset at 1080p delivers 89.3 FPS, which is only 1.3 FPS higher than High, a negligible difference. Given that the CPU is the bottleneck, the data suggests that players should prioritize the High preset at 1080p for a better visual experience without a significant frame rate penalty compared to Ultra, since both sit in the high-80s range.

At 1440p, the pattern shifts slightly. Low delivers 95.2 FPS, Medium drops to 88.7 FPS, and High falls to 79 FPS. The Ultra preset at 1440p produces 75.6 FPS. Here, the Medium preset appears to be the sweet spot for a balance between visual quality and frame rate, as it stays above 85 FPS while providing a noticeable upgrade over Low. The High preset still offers playable performance, but the drop from Medium to High (9.7 FPS) is more substantial than the drop from High to Ultra (3.4 FPS). For a smooth experience, Medium at 1440p is the recommended setting based on the measured data.

At 4K, the situation is more constrained. The Low preset achieves 83.8 FPS, which is the only setting above 80 FPS at this resolution. Medium drops to 73.4 FPS, High to 64.7 FPS, and Ultra to 67.8 FPS. Notably, Ultra is actually slightly faster than High at 4K, a 3.1 FPS difference that is within the margin of measurement error. The data indicates that for 4K, the Low preset is the only one that keeps the frame rate comfortably above 80 FPS, while Medium is still playable. Given the CPU bottleneck, raising settings to High or Ultra at 4K yields diminishing returns, as the frame rate drops below 70 FPS. The best experience per the measured rows would be to use Low at 4K for maximum fluidity, or Medium if visual fidelity is prioritized and a 73 FPS average is acceptable.

FAQ

Q: Why does the RTX 5090 not produce higher frame rates in Thief?

A: The data shows that the AMD Ryzen 5 2600 is the limiting component. The frame rate differences between 1080p and 4K are relatively small (e.g., 110.1 FPS vs. 83.8 FPS at Low settings), which indicates the CPU cannot keep up with the GPU, even at lower resolutions.

Q: What is the best setting for 1080p gaming with this combo?

A: Based on the measured FPS, the High preset at 88 FPS is recommended over Ultra at 89.3 FPS, as the visual gain from High to Ultra is minimal while the performance is nearly identical. Low at 110.1 FPS is an option for competitive play, but High offers a better visual experience for a modest frame rate cost.

Q: Is the RTX 5090's performance in line with its benchmark scores?

A: No. The RTX 5090 ranks in the 94th percentile of all GPUs with a passmark G3D score of 39650, but in Thief, its performance is capped by the CPU. The measured frame rates are far below what the GPU is capable of, as evidenced by the muted resolution scaling.

Q: How does the Ryzen 5 2600 compare to its nearest rivals in terms of CPU benchmarks?

A: The Ryzen 5 2600 has an average benchmark score of 19616, which is within 0.7% of the Intel Core i7-10700F (19485) and 0.6% of the AMD Ryzen 5 7530U (19508). It is slightly behind the Intel Core i5-1345U (19691) by 0.4%.

Q: Does the Ultra preset ever outperform High in this game?

A: Yes, at 4K resolution, the Ultra preset delivers 67.8 FPS, which is 3.1 FPS higher than the High preset's 64.7 FPS. This is an anomaly likely due to the CPU bottleneck, where the specific workload distribution of the Ultra preset is marginally less taxing on the CPU.

Q: What is the combo's rank among all tested combinations?

A: This specific pairing ranks 993rd out of 1642 tested combos in Thief, placing it in the lower-middle portion of the database, which reflects the CPU limitation.

Measured FPS Breakdown

The complete dataset provides a detailed view of performance across all resolution and settings combinations. At 1920x1080, the Low preset leads with 110.1 FPS, followed by Medium at 97.7 FPS, Ultra at 89.3 FPS, and High at 88 FPS. The spread from Low to High is 22.1 FPS, which is significant but shows that even the lowest setting does not reach the 120 FPS mark, further confirming the CPU bottleneck.

At 2560x1440, the order remains consistent: Low at 95.2 FPS, Medium at 88.7 FPS, High at 79 FPS, and Ultra at 75.6 FPS. The drop from Low to Ultra is 19.6 FPS, which is a similar relative gap to 1080p. The data indicates that at 1440p, the CPU constraint becomes even more pronounced, as the frame rates are only slightly lower than 1080p despite the higher pixel count.

At 3840x2160, the Low preset is again the top performer at 83.8 FPS, followed by Medium at 73.4 FPS, Ultra at 67.8 FPS, and High at 64.7 FPS. The most striking finding is that the Ultra preset is faster than High at this resolution, which is unusual and suggests that the High preset may trigger a specific CPU workload that the Ultra preset does not. The overall trend across all resolutions is that the frame rate is primarily determined by the CPU's ability to process the game logic, not by the GPU's rendering capacity.

How This Combo Ranks

In the global database of tested hardware combinations for Thief, this pairing of the AMD Ryzen 5 2600 and NVIDIA GeForce RTX 5090 ranks 993rd out of 1642 combos. This places it in the 60th percentile of all tested systems, meaning it outperforms about 40% of combinations but lags behind the majority. The rank is notably low for a GPU of this caliber, as the RTX 5090 is in the 94th percentile of all GPUs. This discrepancy highlights the impact of the CPU on the overall result.

The CPU's rank among all processors is 77th percentile, which is respectable for a 2018 chip, but its average benchmark score of 19616 is only slightly above the Intel Core i5-1345U (19691, 0.4% faster) and the Intel Core i5-11600KF (19740, 0.6% faster). However, these rivals are not necessarily indicative of gaming performance. The combo rank of 993 suggests that many other CPU-GPU pairings achieve higher frame rates in Thief, likely because they feature more modern CPUs that can better feed high-end GPUs. The data clearly shows that the RTX 5090 is wasted in this configuration, as the CPU becomes the limiting factor at every resolution and settings level.

CPU Role

The AMD Ryzen 5 2600 is a 6-core, 12-thread processor based on the Zen architecture, with a base clock of 3.40 GHz and a boost clock of 3.90 GHz. It has a 16 MB shared L3 cache and supports DDR4 memory with a dual-channel bus. In terms of raw benchmarks, it scores 11186 in Cinebench R23 multi-core and 1579 in single-core, placing it in the 77th percentile of all CPUs. Its nearest rivals include the Intel Core i5-1345U and AMD Ryzen 5 7530U, with performance deltas of less than 1%.

In Thief, the CPU's role is critical, as the game appears to be highly dependent on single-thread performance. The measured FPS data shows that even at 4K with Low settings, the frame rate is capped at 83.8 FPS, which is far below what the RTX 5090 could deliver if it were unconstrained. The CPU's single-core Cinebench R23 score of 1579 is modest by modern standards, and this likely explains why the frame rates plateau in the 60-110 FPS range. The boost clock of 3.90 GHz is relatively low for a gaming CPU, and the older Zen architecture does not have the per-core efficiency of newer designs.

The data suggests that the Ryzen 5 2600's 6 cores are sufficient for Thief's threading model, but the clock speed and IPC limits are the binding constraint. The fact that the Ultra preset at 4K (67.8 FPS) outperforms High (64.7 FPS) indicates that the CPU's workload varies with specific settings, but the overall ceiling remains. For this game, the CPU is the clear bottleneck, and upgrading to a processor with higher single-thread performance would yield significant frame rate gains, even with the same RTX 5090.