Thief
This combination provides smooth gameplay with an average of 96 FPS, suitable for most gaming scenarios.
Average FPS: resolution vs quality settings
| Ultra | High | Medium | Low | |
|---|---|---|---|---|
| 4K Ultra HD | 74 | 75 | 83 | 92 |
| 1440p QHD | 89 | 90 | 99 | 110 |
| 1080p Full HD | 100 | 102 | 113 | 125 |
Cell color: green is 120+ FPS, teal is 60+, amber is 30+, red is below 30.
Thief with Intel Core i5-10400F + NVIDIA GeForce RTX 5090 D, In-Depth Analysis
The pairing of the Intel Core i5-10400F with the NVIDIA GeForce RTX 5090 D in Thief presents a stark example of component bottlenecking in a legacy title. The benchmark data shows a system where the GPU’s immense raw capability is consistently restrained by the CPU’s architectural limits, particularly at lower resolutions. This analysis walks through the measured FPS figures to explain how resolution scaling, component roles, and settings interact to produce the final performance profile, which ranks in the 44th percentile of tested combinations for this game.
Resolution Scaling
The measured FPS data reveals an unusual scaling pattern across resolutions. At 1080p with High settings, the combo delivers 102 FPS. Moving to 1440p High, performance drops to 90 FPS, and at 4K High, it falls further to 75 FPS. This represents a 26.5% decrease from 1080p to 4K. However, the more telling observation is the absolute frame rates themselves. A GPU with the RTX 5090 D’s specifications should theoretically produce much higher frame rates in a 2014 title, yet the data shows it capped at 125 FPS at 1080p Low. The relatively small gap between Low and Ultra settings at each resolution—for instance, only 25 FPS separating Low (125 FPS) from Ultra (100 FPS) at 1080p—suggests that the CPU is the limiting factor.
The scaling from 1440p to 4K shows a more pronounced drop. At Medium settings, the FPS goes from 99 at 1440p to 83 at 4K, a 16.2% reduction. This larger drop at higher resolution indicates that the GPU begins to take on more of the workload as pixel count increases, but it never reaches a state where it is fully saturated. The data points to a scenario where the RTX 5090 D is idling much of the time, waiting on the i5-10400F to feed it draw calls. The fact that 4K Ultra still manages 74 FPS, only 1 FPS less than 4K High, reinforces that the GPU has ample headroom. In a GPU-bound scenario, the gap between High and Ultra at 4K would typically be larger. Here, the difference is negligible, which is a hallmark of a CPU-limited configuration.
GPU Role
The NVIDIA GeForce RTX 5090 D brings a formidable set of specifications to this test. Its 32 GB of GDDR7 memory on a 512-bit bus provides 1.79 TB/s of bandwidth, which is far beyond what a game like Thief requires. The GPU’s compute capabilities are equally massive, with 21,760 shading units and a boost clock of 2407 MHz. In terms of raw throughput, the data lists 104.8 TFLOPS for FP32 operations. These numbers place the GPU in the 92nd percentile of all GPUs, with an average benchmark score of 77,712.
Despite this hardware, the measured FPS in Thief does not reflect the GPU’s potential. The passmark G3D score of 44,065 and the 3DMark Steel Nomad DX12 score of 14,326 show a card that excels in synthetic benchmarks. However, in this game, the GPU’s role is diminished. The VRAM capacity is effectively irrelevant; 32 GB cannot be fully utilized by a title from 2014. The memory clock of 1750 MHz (28 Gbps effective) and the resulting bandwidth are likewise underutilized. The data suggests the GPU is operating well below its capacity. The nearest rivals in the GPU benchmark list, such as the AMD Radeon RX 6650M XT (deltaPct 1.1%) and the NVIDIA Tesla P100 PCIe 16 GB (deltaPct -2.4%), have significantly lower specifications, yet the performance difference in Thief would likely be minimal due to the CPU bottleneck. The GPU’s 176 ROPs and 680 TMUs are more than sufficient for this game’s rendering demands, but they are never fully engaged.
CPU Role
The Intel Core i5-10400F is a 6-core, 12-thread processor from the Comet Lake architecture, running on the Intel Socket 1200 platform. Its base clock is 2.90 GHz with a boost clock of 4.30 GHz. The CPU has 12 MB of shared L3 cache. In synthetic benchmarks, it scores 10,283 in Cinebench R23 multi-core and 1,451 in single-core. Its average benchmark score is 14,185, placing it in the 68th percentile of all CPUs. The nearest CPU rivals include the Intel Xeon 6756E (deltaPct 0.2%) and the AMD Ryzen 3 7320C (deltaPct -0.6%), indicating that the i5-10400F sits in a mid-range performance tier.
In Thief, the CPU’s role is critical. The game’s engine, which is a stealth-focused title from 2014, likely relies heavily on a few threads for game logic and draw call submission. The i5-10400F’s single-thread performance, as indicated by a Geekbench single-core score of 1,420 and a Passmark single-thread score of 2,541, becomes the primary constraint. The data shows that at 1080p Low, the FPS is 125. This is likely the ceiling for this CPU in this game. The GPU could render frames faster, but the CPU cannot issue instructions quickly enough. The 4.30 GHz boost clock is decent, but the architecture is older, and the IPC (instructions per clock) is lower than newer designs. The fact that the FPS does not scale down significantly from Low to Ultra at 1080p (125 to 100 FPS) confirms that the CPU is the limiting component. The GPU has enough headroom to handle the increased settings without impacting frame time, but the CPU’s throughput on the main thread is the bottleneck.
Settings Recommendations
Based on the measured rows, the optimal settings for this combo are not about maximizing GPU load but about finding the sweet spot where the CPU bottleneck is least impactful. At 1080p, the Ultra preset yields 100 FPS, while High yields 102 FPS. The difference is minimal, so choosing Ultra provides better visual quality for the same playable frame rate. However, the Medium preset at 1080p offers 113 FPS with a minimum of 96 FPS and a maximum of 130 FPS. This represents a 13% performance increase over Ultra with likely minor visual differences. For competitive or high-refresh-rate play, Medium at 1080p is the most balanced choice, as it offers a higher average and a minimum frame rate that stays above 60 FPS.
At 1440p, the pattern is similar. Ultra provides 89 FPS, while Medium provides 99 FPS with a minimum of 85 FPS. The Medium preset is again the recommended option, as it provides a 10 FPS advantage over Ultra while remaining well above 60 FPS. At 4K, the differences narrow. Ultra offers 74 FPS, and Medium offers 83 FPS with a minimum of 71 FPS. For a smoother experience at 4K, Medium is the better choice, as it keeps the minimum FPS above 70. The Low preset at 4K gives 92 FPS, but the visual downgrade is substantial. Given that the GPU has ample headroom, the recommendation is to prioritize image quality by using Medium or High settings. Specifically, Medium is the best overall preset across all resolutions, as it consistently provides the highest FPS above the Ultra preset with a minimum frame rate that is acceptable for smooth gameplay. High is a close second, but the data shows Medium has a 10-11 FPS advantage at 1080p and 1440p.
How This Combo Ranks
The combo rank for this pairing in Thief is 1,261 out of 2,851 tested combinations. This places it in the middle of the pack, which is surprising given the high-end GPU. The data suggests that many other combinations, likely those with more balanced CPUs and GPUs, achieve higher frame rates in this game. The 44th percentile rank indicates that over half of the tested combos perform better. This is a direct consequence of the CPU bottleneck. A system with a lower-end GPU but a stronger single-threaded CPU would likely rank higher in this specific game.
The GPU’s percentile rank of 92 is excellent, but it does not translate to a high combo rank because the CPU’s percentile rank of 68 is not high enough to keep up. The delta between the GPU’s capability and the CPU’s capability is the defining factor of this combo’s performance. The measured FPS data shows that the system performs at a level typical of a mid-range CPU pairing, despite the presence of a flagship GPU. This rank serves as a cautionary tale about pairing a 10th-gen i5 with a 50-series GPU for older titles. The GPU is effectively wasted in this configuration for Thief, as the CPU cannot feed it enough data to differentiate it from a much cheaper graphics card.
FAQ
Q: Why does the RTX 5090 D not produce higher frame rates in Thief?
A: The benchmark data indicates the Intel Core i5-10400F is the limiting component. The FPS does not scale significantly between Low and Ultra settings, particularly at 1080p, which is a clear sign of a CPU bottleneck. The GPU’s massive compute resources are underutilized.
Q: Is 4K gaming viable with this combo in Thief?
A: Yes, 4K is viable. The measured data shows 4K Ultra runs at 74 FPS and 4K Medium at 83 FPS. Both are above 60 FPS, so the experience is smooth. The CPU bottleneck is less pronounced at this resolution because the GPU has more work to do.
Q: What is the best settings preset to use?
A: The Medium preset offers the best balance. At 1080p, it delivers 113 FPS; at 1440p, 99 FPS; and at 4K, 83 FPS. It consistently outperforms the Ultra preset by a margin of 9-13 FPS while providing a minimum frame rate above 70 in all cases.
Q: How does this combo rank compared to others?
A: This combo ranks 1,261st out of 2,851 tested combinations, placing it in the 44th percentile. This indicates that the pairing is not optimal for this game, primarily due to the CPU’s inability to keep pace with the GPU.
Q: Does the 32 GB of VRAM matter for this game?
A: No. Thief is a 2014 title, and its memory requirements are minimal. The 32 GB of GDDR7 memory and the 1.79 TB/s bandwidth are far more than the game can use. The VRAM has no impact on the measured FPS results.
Q: Is the CPU upgrade the only way to improve performance?
A: Based on the data, yes. The FPS is capped by the CPU’s single-thread performance. Upgrading to a CPU with higher single-core speed would likely yield higher frame rates, as the GPU has ample headroom to render more frames.
Measured FPS Breakdown
The exact measured FPS figures are as follows. At 1920x1080, the Low preset yields an average of 125 FPS. The Medium preset yields an average of 113 FPS, with a minimum of 96 FPS and a maximum of 130 FPS. The High preset yields 102 FPS, and the Ultra preset yields 100 FPS. At 2560x1440, Low yields 110 FPS. Medium yields 99 FPS, with a minimum of 85 FPS and a maximum of 114 FPS. High yields 90 FPS, and Ultra yields 89 FPS. At 3840x2160, Low yields 92 FPS. Medium yields 83 FPS, with a minimum of 71 FPS and a maximum of 96 FPS. High yields 75 FPS, and Ultra yields 74 FPS. The data shows a consistent decline from Low to Ultra across all resolutions, but the magnitude of the decline is small, which reinforces the conclusion that the CPU is the performance limiter. The spread between Low and Ultra is 25 FPS at 1080p, 21 FPS at 1440p, and 18 FPS at 4K. This decreasing spread with increasing resolution further confirms that the GPU becomes more involved at higher pixel counts, but never becomes the primary bottleneck.
Hardware Specifications
Intel Core i5-10400F
NVIDIA GeForce RTX 5090 D
FPS Benchmarks by Resolution & Settings
Performance data for Intel Core i5-10400F + NVIDIA GeForce RTX 5090 D in Thief
| Resolution | Settings | Avg FPS |
|---|---|---|
| 3840x2160 | Low | 92.0 |
| 3840x2160 | Medium | 83.0 |
| 3840x2160 | High | 75.0 |
| 3840x2160 | Ultra | 74.0 |
| 2560x1440 | Low | 110.0 |
| 2560x1440 | Medium | 99.0 |
| 2560x1440 | High | 90.0 |
| 2560x1440 | Ultra | 89.0 |
| 1920x1080 | Low | 125.0 |
| 1920x1080 | Medium | 113.0 |
| 1920x1080 | High | 102.0 |
| 1920x1080 | Ultra | 100.0 |
Thief with ii5-10400F + Other GPUs
See how Thief performs with the same CPU but different graphics cards.
Thief with RTX 5090 D + Other CPUs
See how Thief performs with the same GPU but different processors.
Other Games with ii5-10400F + RTX 5090 D
Explore FPS benchmarks for other games using this same hardware combination.