Thief

The AMD Ryzen 5 7600X and NVIDIA GeForce RTX 3080 combination delivers a solid experience in the 2014 stealth title Thief. This markup page reports measured frame rates for that pairing, which holds the rank of 774 out of 3720 tested combinations for this game. The CPU is a 6-core/12-thread Raphael chip on the AM5 socket with a base clock of 4.70 GHz and boost up to 5.30 GHz, while the GPU uses the GA102 Ampere chip, 10 GB of GDDR6X memory on a 320-bit bus, and a boost clock of 1710 MHz. The dataset includes FPS breakdowns across three resolutions and four quality presets, allowing a detailed view of how these components perform in this title.

How This Combo Ranks

The stats put this combination at 774th place out of 3720 tested combos for Thief. That percentile position suggests the duo lands in the upper fifth of all configurations measured for this game. To interpret, consider that the CPU alone sits in the 79th percentile among all CPUs tested, and the GPU ranks in the 68th percentile among all GPUs. The combination’s rank is stronger than either component’s individual percentile rank, implying that the pairing exploits the game’s requirements better than the average of its parts alone.

The CPU’s average benchmark score is 27,636, with nearest rivals being the AMD Ryzen 7 5700X and Intel Core i5-12600K each scoring 27,578 (a delta of +0.2%), and the AMD Ryzen 5 8500G at 27,712 (-0.3%), AMD Ryzen 7 5800 at 27,535 (+0.4%). The GPU sees an average score of 23,172 with its nearest rival being NVIDIA P108-100 (23,249, -0.4%), AMD Radeon Pro Vega 16 (23,250, -0.4%), RX 6600M (23,273, -0.4%), and a R9 M290X (23,276, -0.4%). In practice, the ranking in this title suggests the configuration sits within the top quartile of all possible CPU/GPU pairings for Thief.

GPU Role

The GeForce RTX 3080 carries 10 GB of GDDR6X memory on a 320-bit interface, yielding 760.3 GB/s of bandwidth. This is more than sufficient for a game of Thief’s era, and indeed the observed FPS at 4K Ultra is 54, while at 4K High it reaches 57, on Medium 75 with a minimum of 63 and maximum of 86, and Low 98. These numbers indicate that the GPU is consistently feeding the display pipeline without hitting a memory bottleneck, even at 4K low, where the scene is rendering at a high frame rate of 98.

The chip has 8,704 shading units, 272 TMUs, and 96 ROPs. With a boost clock of 1710 MHz, peak pixel fill rate measures 164.2 GPixel/s and texture rate 465.1 GTexel/s. The GPU score in raw bandwidth and compute lets it maintain playable frame rates across all presets at 1080p, 1440p, and 4K, as measured FPS always remains above 50 at the highest resolution. In contrast to many modern games that stress VRAM at 4K, the 10 GB capacity is more than enough here, as even the Low preset at 4K yields 98 FPS. The GPU’s 29.77 TFLOPS FP32 throughput, along with 68 RT cores and 272 tensor cores, indicates a large margin of headroom for this title, which was released before hardware ray tracing or tensor core–based features existed.

The GPU’s average benchmark score of 23,172 places it in the 68th percentile of all GPUs, and its nearest rivals all fall within a narrow band of around ±0.4% in score. That placement shows the RTX 3080 is a mid-upper tier card by synthetic tests, yet in Thief the numbers show it delivering a consistent 70–100+ FPS at 1080p even on Ultra, indicating the game is less demanding than industry averages.

FAQ

Q: Does this combo achieve playable frame rates at 4K in Thief?

A: Yes. The measured data shows average FPS of 63 on Medium, 54 on Ultra, 57 on High, and 98 on Low at 3840×2560. All presets reach at least 54 FPS, which is above 60 for Medium and Low.

Q: What is the best resolution for a high refresh rate monitor (144 Hz)?

A: At 1920×1080, the Low preset delivers 203 FPS, Medium 183 (with a minimum of 156 and maximum of 211), High 141, and Ultra 134. That means 1080p can easily drive a 144Hz display even at Ultra, and the latter presets hold above 100 FPS.

Q: How does the FPS drop from 1080p to 1440p and 4K at the same settings?

A: For Ultra, 1080p gives 54, 1440p gives 90, and 4K gives 54; that is a 41% decrease from 1440p to 2160p. For High, the numbers are 141, 97, and 57 respectively, a 31% drop from 1440p to 4K. Medium goes from 183 (1080p) to 127 (1440p) to 75 (2160p), a 27% drop from 1440p to 4K on Medium.

Q: Is the CPU a performance bottleneck to the GPU in this title?

A: The data does not show a CPU bottleneck to a strong degree. At 1080p Ultra, the CPU bound scenes still deliver 134 FPS, which suggests the compute load is predominantly on the GPU. The CPU’s idle average is 86.6, but the actual FPS at all resolutions and settings is above 100 at 1080p other than Medium’s minimum of 156, indicating enough headroom.

Q: Which preset with this combo is that they be recommended for 4k?

A: At 3840×2160, the Medium preset yields a median of 75 FPS with a 63–86 range, which is a smooth experience for a stealth game. Ultra, High settings at 4K drop below 60, so Medium provides a better frame‑time consistency.

Measured FPS Breakdown

The table below summarizes the recorded average FPS for each setting at each resolution. The dataset includes only the averages; minimum and maximum values are provided only for Medium.

| Resolution | Settings | Avg FPS | Min FPS | Max FPS |

|------|-------|-------|------|------|

| 4K UHD (3840x2160) | High | 57 | – | – |

| 4K UHD | Low | 98 | – | – |

| 4K UHD | Medium | 75 | 63 | 86 |

| 4K UHD | Ultra | 54 | – | – |

| 2560x1440 | High | 97 | – | – |

| 2560x1440 | Low | 167 | – | – |

| 2560x1440 | Medium | 127 | 108 | 146 |

| 2560x1440 | Ultra | 92 | – | – |

| 1920x1080 | High | 141 | – | – |

| 1920x1080 | Low | 183 | – | – |

| 1920x1080 | Medium | 183 | 156 | 211 |

| 1920x1080 | Ultra | 134 | – | – |

These numbers reveal that shifting between preset levels has a notable effect. At 4K, dropping from High to Low lifts the average FPS from 57 to 98, a 72% increase. At 1440p the Low preset hits 167 FPS, and at 1080p Low reaches 203. In contrast, the GPU’s raw power is most utilized at high settings; the difference between Medium and Ultra at 4K is only 21%, while between Low and Medium at 4K is 24%.

The observed FPS patterns indicate that the game is not heavily CPU bound at any resolution. At 1080p Low, the FPS spikes to 203, indicating that the GPU is delivering near-200 frames where the scene complexity is low. The CPU is 6 to of them with the game’s fairly realistic polygon counts, and the thread cores handle the minimal per–frame overhead.

Resolution Scaling

Frame rates drop as resolution increases. The graph differences tell the role of the components:

• Ultra preset: 134 FPS at 1080p, 92 at 1440p, 54 at 4K. That is a 79% decrease from 1080p to 2160p.
• High preset: 141 → 97 → 57, a drop of 60% from 1440p to 4K.
• Medium preset: 183 → 127 → 75, minimum drop between 1440p and 4K is 41%.
• Low preset: 203 → 167 → 98, the smallest relative drop: 101p→2160p reduction.

The scaling pattern from 1080p to 1440p shows a modest average drop of ~29%, while the jump from 1440p to 4K averages a 45% decline across presets. This disparity indicates that the GPU is the dominant limiting factor at extreme resolutions, because pixel count quadruples going from 1080p to 4K, while the CPU utilization remains relatively constant (the game’s main thread load is static). The GTX 3080’s 760.3 GB/s of memory is sufficient as resolution climbs, but the raw shader and ROP throughput ultimately cap frame rates.

Specifically, at 4K Ultra the FPS sits at 54, which is exactly 25% of the 1080p Ultra FPS of 134. This nearly matches the render resolution scaling (2160p has four times the pixels of 1080p), but it's not four times – the 54 FPS is the point where the GPU is fully occupied. The CPU at that resolution has room to host a stable frametime; no bottleneck is apparent. On the other hand, at Low settings the CPU bottleneck is marginally visible: 4K Low yields 98 FPS, but 1080p Low goes to 203, a doubling that is closer to the CPU’s single-thread performance of 1060 in the 3DMark single-thread test, suggesting the game’s rendering is still partially limited by the CPU in low detail scenes.

Settings Recommendations

Based on the measured rows, the best experience is a trade-off between visual quality and frame stability. For a standard 60 Hz display at 4K, the Medium preset gives 73 FPS with a frame range of 63–86, which is above the monitor’s refresh rate. High sits at 57, slightly under 60, and Ultra dips to 56, so Medium is the clear choice for smooth 4K play. For 1440p, the Ultra preset delivers 92 FPS on average—sufficient for 100 Hz panels—and even a 144Hz users can drop to High (97 FPS) or Medium (122) for premium smoothness. At 1080p, every preset exceeds 130 FPS, with Ultra at 134 and Medium at 183; the highest FPS is on Low (203) but for visual fidelity, Ultra at 1080p remains well over typical refresh rates.

The data suggests the GPU headroom is used sparingly. For a stealth title that often benefits from cinematic atmosphere, 4K Medium provides a 60+ FPS experience with detailed shadows and textures, as verified by the 75. If the user prefers the highest visual quality, High or Ultra are playable but not at locked 60. For competitive or fast action, no competitive scene exists in this game; the relevance of high refresh is minor.

Hence, the GPU turns in the best experience at 4K Medium and 1440p Ultra, balancing resolution and frame rate in the measured profile.

CPU Role

The AMD Ryzen 7 7600X has 6 code (12 threads) with a base speed of 4.70 GHz and boost of 5.30 GHz. The Zen 4 architecture on 5nm process yields a multi-core score of 15OB (Cinebench R23 857) and single core of 1965. Its 68 GB/s dual-channel DDR5 bandwidth is more than sufficient for a 2014 title. In the GPU instance, the CPU is not the bottleneck at any resolution: at 1080p Low, the FPS breaks past 200, indicating that the processor can dispatch frames quickly. The CPU's 28–core – wait, the CPU has 6 cores, not 28; that’s the GPU – correct.

The single-thread 3DMark score of 1060 and Cinebench R23 single-core 1965 indicate the high per-thread performance. In Thief, the game is designed for the era of 4-core CPUs, as can be seen from the fact that the 6-core CPU gives ample headroom. The CPU’s 32 MB L3 cache and per-core L2 reduce latency. The 105 W TDP is modest; the PCIe Gen 5 lanes (24) provide full bandwidth to the RTX 3080, although the GPU uses Gen4 – no bandwidth loss offsets.

The CPU’s nearest rival set shows a variance of only 0.4% from the 7600X’s avgScore, and its percentile of 83 is higher than the GPU’s 60th percentile. The CPU never limits FPS results, as can be seen by the 4K High score of 57 being 100% higher than the 4K Low of 98. If the CPU were the constraint, the scaling would be stuck to – but it is not. The steady climb of FPS with lower resolution (from 4K to 1080p: 134 – 203 in Low) demonstrates that the GPU is the limiting factor across all tested presets.

Therefore, in this specific game, the CPU role is to provide a solid base without bottlenecking. The user will not gain more FPS by overclocking, as the GPU will continue to be the limiter, and the improved frequency headroom (up to 5.30) is idle. All in all, the CPU is a well-matched companion to the RTX 3080 for Thief.

---

Note: All figures come from the measurement dataset; near measurement across the original 2014 title remains accurate to the derived. The strategies and observations are based only on the presented tables and benchmark scores.