Thief

Thief, the 2014 stealth classic, is a fascinating case study for modern hardware. The data for this combination of an Intel Core i9-14900K and an NVIDIA GeForce RTX 3080 reveals a title that scales dramatically with settings and resolution, but with performance characteristics that suggest a game engine from a bygone era. The measured results place this combo in the 692nd position out of 3720 tested combinations, putting it in the top 20% of all systems tested, a strong showing for a decade-old game.

Settings Recommendations

The measured data across the three resolutions and four quality presets paints a clear picture of how this hardware configuration interacts with Thief's engine. The most striking observation is the massive performance headroom available at lower settings, which suggests that the Ultra and High presets are the only ones that meaningfully stress the GPU. The data shows a consistent pattern where moving from Low to Medium incurs a significant performance penalty, but the jump from Medium to High, and especially from High to Ultra, is relatively smaller in percentage terms.

For players seeking the best experience per the measured rows, the High preset at 2560x1440 appears to be the sweet spot. It delivers an average of 97 FPS, which is well above the 60 FPS threshold for smooth gameplay, while providing a substantial visual upgrade over the Medium preset's 127 FPS. The jump from Medium to High at 1440p represents a 30 FPS cost for better fidelity, which is a reasonable trade-off. Pushing to Ultra at this resolution drops the average to 92 FPS, a further 5 FPS reduction for what is often an imperceptible visual gain in a game of this era.

At 1920x1080, the High preset produces an average of 141 FPS, which is excellent for high-refresh-rate monitors. The Ultra preset still manages 134 FPS, making it a viable option for those who prioritize maximum visual quality over a few extra frames. However, the data suggests that the Medium preset at 1080p, with its 185 FPS average, is unnecessary unless targeting a very high refresh rate display, as the visual sacrifice is not worth the performance gain when High is already so fast.

The 4K resolution data tells a different story. At 3840x2160, the High preset averages 57 FPS, which is borderline for smooth gameplay. The Ultra preset drops to 54 FPS, and while these are playable, they are not ideal for a stealth game where quick reactions can be necessary. The Medium preset at 4K, averaging 75 FPS, is the best choice for a high-fidelity experience at this resolution, offering a comfortable margin above 60 FPS. The Low preset at 4K, at 98 FPS, seems like a waste of the visual capabilities of the hardware, as the GPU is clearly not the limiting factor at that setting.

GPU Role

The NVIDIA GeForce RTX 3080 is the primary driver of performance in Thief at higher settings. The GPU's 10 GB of GDDR6X memory on a 320-bit bus, providing 760.3 GB/s of bandwidth, is more than sufficient for this game's textures and geometry. The measured data shows that the GPU's role becomes more pronounced as resolution and settings increase, which is the expected behavior for a graphics-bound title.

The RTX 3080's boost clock of 1710 MHz, based on the Ampere architecture with 8704 shading units, delivers the raw compute necessary to push the pixel rates seen in the results. The data shows a clear scaling pattern from 1080p to 4K that is consistent with GPU-bound performance. At 1080p High, the average is 141 FPS, which drops to 97 FPS at 1440p, and then to 57 FPS at 4K. This progression indicates that the GPU is being utilized more heavily at higher resolutions, as the frame rate drops proportionally with the increase in pixel count.

Interestingly, the GPU's performance ceiling is not reached at lower settings. At 1080p Low, the system achieves 230 FPS, which is likely hitting a CPU or engine limit rather than a GPU limit. The RTX 3080's Passmark G3D score of 25086 places it in the 68th percentile of all GPUs, and its performance in Thief shows that it is a capable card for this title, though not overwhelmingly so at 4K Ultra. The 57 FPS average at 4K High suggests that while the GPU is powerful, Thief's engine is not optimized for the highest resolutions, and the card's 10 GB VRAM is not being taxed by a 2014 release.

How This Combo Ranks

The combination of the Intel Core i9-14900K and the NVIDIA GeForce RTX 3080 achieves a rank of 692 out of 3720 tested combinations in Thief. This places the system in the top 18.6% of all tested configurations, which is a strong result for a game that is not particularly demanding by modern standards. The rank suggests that while this hardware is high-end, there are many other combinations that perform similarly or better in this specific title.

The CPU's percentile rank of 95 among all processors indicates that the i9-14900K is a top-tier chip, but its performance in Thief is not the deciding factor for the combo's rank. The GPU's percentile rank of 68 is more reflective of the system's overall standing, as Thief is a GPU-bound title at higher settings. The data shows that the combo's rank is likely influenced more by the GPU's performance than the CPU's, given that the game does not heavily utilize the CPU's 24 cores and 32 threads.

When comparing to other tested combos, the i9-14900K's nearest rivals in synthetic benchmarks, such as the Intel Core i9-14900KF with a delta of -0.3%, perform nearly identically, which suggests that the CPU choice is not a significant differentiator in Thief. The GPU's nearest rivals, including the NVIDIA P106-100 with a delta of -0.3%, are far less powerful cards, yet the RTX 3080's lead in this game is not as pronounced as the synthetic benchmarks might suggest, due to the game's age and engine limitations.

Measured FPS Breakdown

At 1920x1080, the performance is exceptionally high across all settings. The Low preset achieves an average of 230 FPS, which is likely a CPU or engine cap. The Medium preset averages 185 FPS, with a minimum of 157 and a maximum of 213 FPS. The High preset drops to 141 FPS, and the Ultra preset averages 134 FPS. The data shows a clear step-down in performance as settings increase, with the most significant drop occurring between Medium and High, a difference of 44 FPS.

Moving to 2560x1440, the performance scales down accordingly. The Low preset averages 167 FPS, a 63 FPS drop from 1080p, which is a 27% reduction. The Medium preset averages 127 FPS, with a minimum of 108 and a maximum of 146 FPS. The High preset averages 97 FPS, and the Ultra preset averages 92 FPS. The gap between Medium and High at 1440p is 30 FPS, which is a smaller absolute difference than at 1080p but similar in percentage terms.

At 3840x2160, the performance becomes more demanding for the GPU. The Low preset averages 98 FPS, which is still playable but a significant drop from the lower resolutions. The Medium preset averages 75 FPS, with a minimum of 63 and a maximum of 86 FPS. The High preset averages 57 FPS, and the Ultra preset averages 54 FPS. The data shows that at 4K, the difference between High and Ultra is only 3 FPS, suggesting that the Ultra preset is not well-optimized or that the GPU is hitting a performance wall.

Resolution Scaling

The scaling from 1080p to 4K reveals important information about the limiting component in this system. Moving from 1920x1080 to 2560x1440 at High settings, the average FPS drops from 141 to 97, a 31% reduction. This is a larger percentage drop than the pixel count increase of 78%, which is unusual and suggests that the game engine is not scaling efficiently with resolution. The drop from 1440p to 4K at High settings is from 97 to 57 FPS, a 41% reduction, which is closer to the expected scaling for a GPU-bound workload.

The Low preset scaling is even more revealing. At 1080p Low, the system achieves 230 FPS, but at 1440p Low, it drops to 167 FPS, a 27% reduction. At 4K Low, it drops to 98 FPS, a 41% reduction from 1440p. This pattern suggests that even at Low settings, the GPU is becoming the limiting factor at higher resolutions, but the CPU or engine is still holding back performance at 1080p. The fact that 4K Low (98 FPS) is much closer to 4K Medium (75 FPS) than 1080p Low (230 FPS) is to 1080p Medium (185 FPS) indicates that the GPU is being saturated at 4K.

The data implies that the RTX 3080 is the primary bottleneck at 4K, while at 1080p, the game's engine or the CPU's single-threaded performance may be limiting the frame rate. The i9-14900K's single-core performance, as measured by a Cinebench R23 single-core score of 2262.5, is excellent, but Thief's engine may not be able to fully utilize it at lower resolutions.

CPU Role

The Intel Core i9-14900K is a 24-core, 32-thread processor with a base clock of 3.20 GHz and a boost clock of 6.00 GHz. Its architecture, Raptor Lake, is built on a 10 nm process. In Thief, a game from 2014, the CPU's role is expected to be minimal, as the game was designed for far less powerful processors. The data confirms this, as the CPU's massive multi-threading capabilities are not reflected in the frame rates at higher resolutions.

The CPU's Passmark multi-thread score of 58674 is in the 95th percentile of all CPUs, and its average benchmark score of 79097 is within 1.7% of the AMD Ryzen AI Max+ 395 and 1.2% of the AMD EPYC 7413. However, in Thief, the difference between the i9-14900K and these rivals would be negligible. The game likely relies on a few cores, and the i9-14900K's single-core performance, with a Passmark single-thread score of 4697, is more than sufficient.

The data shows that at 1080p Low, the system achieves 230 FPS, which is likely near the engine's maximum frame rate cap or a CPU limitation. The i9-14900K's boost clock of 6.00 GHz is exceptionally high, but Thief's engine may not be able to take advantage of it fully. The CPU's role becomes even less significant at 4K, where the GPU is clearly the bottleneck, as evidenced by the 54 FPS average at Ultra settings. The CPU is not a limiting factor in this game, and its high core count is irrelevant for Thief's workload.

FAQ

Q: What is the best resolution and settings combination for a smooth experience?

A: The data indicates that 2560x1440 with High settings, averaging 97 FPS, is the best balance of visual quality and performance. At 4K, the Medium preset at 75 FPS is recommended over High at 57 FPS, as the visual difference is minimal but the performance gain is significant.

Q: Is the RTX 3080 sufficient for 4K gaming in Thief?

A: Based on the measured data, the RTX 3080 can handle 4K, but not at the highest settings. At 4K Ultra, the average FPS is 54, which is playable but not ideal. The High preset at 57 FPS is similar, while Medium at 75 FPS offers a much better experience. The GPU is the limiting factor at this resolution.

Q: Why is the FPS so high at 1080p Low, and what limits it?

A: The 230 FPS average at 1080p Low is likely limited by the game engine or CPU, rather than the GPU. The RTX 3080 is far more powerful than needed for this setting, and the i9-14900K's high single-core performance is not fully utilized by the game's engine, which was built for 2014 hardware.

Q: How does this combo compare to other systems in Thief?

A: This combination ranks 692nd out of 3720 tested combos, placing it in the top 18.6%. This is a strong result, but the rank is more influenced by the GPU's 68th percentile standing than the CPU's 95th percentile, as Thief is a GPU-bound title at higher settings.

Q: Does the i9-14900K's 24-core design benefit Thief?

A: No, the data suggests that Thief does not utilize the CPU's multi-threading capabilities. The game is from 2014 and likely uses only a few cores. The CPU's single-core performance is more relevant, and its 6.00 GHz boost clock is more than adequate for this title.

Q: What is the performance difference between the High and Ultra presets?

A: The difference is relatively small. At 1080p, High averages 141 FPS and Ultra averages 134 FPS, a difference of 7 FPS. At 1440p, High averages 97 FPS and Ultra averages 92 FPS, a difference of 5 FPS. At 4K, High averages 57 FPS and Ultra averages 54 FPS, a difference of only 3 FPS, making Ultra a poor trade-off for the extra visual load.