Thief

AMD Ryzen 5 3600 and AMD Radeon RX 9070 XT form a pairing that shows a distinct split personality in Thief, depending on where you point the resolution dial. The CPU is a 6-core, 12-thread Zen 2 part from the 3000 series, running a base clock of 3.60 GHz and a boost clock of 4.20 GHz. It has a 32 MB shared L3 cache and supports dual-channel DDR4 memory. In synthetic benchmarks, it scores 15,045 in Cinebench R23 multi-core and 2,124 in single-core, placing it in the 76th percentile of all CPUs. Its average benchmark score of 18,129 is effectively a tie with its closest rivals: the Intel Core i5-11400 sits at 18,150 with a -0.1% delta, while the AMD Ryzen 5 7535HS is at 18,101 with a +0.2% delta.

The GPU, meanwhile, is a RDNA 4.0 part built on the Navi 48 chip at a 4 nm process node. It carries 16 GB of GDDR6 memory on a 256-bit bus, with bandwidth listed at 644.6 GB/s. The boost clock reaches 2970 MHz, and the card has 4096 shading units. It scores 26,795 in Passmark G3D and 72,60 in 3DMark Steel Nomad DX12, putting it in the 82nd percentile of all GPUs. The data shows the CPU is a solid mid-range workhorse, but in a 2014 stealth title like Thief, the GPU is the dominant factor in nearly every measured scenario.

CPU Role — cores, clocks, and how they relate to this game's results

The Ryzen 5 3600 provides six physical cores with simultaneous multithreading, yielding 12 threads. Its base clock of 3.60 GHz and boost of 4.20 GHz are modest by modern standards, but Thief is not a heavily threaded title. The game was released in early 2014, before multi-core scaling became a priority for many developers. Benchmark results indicate the CPU's single-thread performance, reflected in a 3DMark single-thread score of 696 and a Cinebench R23 single-core score of 2,124, is more than sufficient to feed the GPU at 1080p. At 1920x1080 with Low settings, the combo hits 128.3 FPS average, which suggests the CPU is not the primary limiter even at the lowest graphical load.

The CPU's 3DMark 2-thread score of 1,361 and 4-thread score of 2,579 show that even lightly threaded workloads are handled capably. Since Thief's engine likely relies on a few main threads, the 3600's per-core efficiency matters more than its total core count. The L3 cache of 32 MB is generous for the era and helps with game data locality. With a Passmark single-thread score of 2,562, the processor sits comfortably ahead of older 6-core parts like the AMD Ryzen 5 1600, which trails by -0.2% in average score (18,158 vs 18,129). The data shows no scenario in the measured rows where the CPU becomes the bottleneck; even at 1080p Low, FPS remains high, and the GPU still has headroom.

How This Combo Ranks — use comboRankInGame vs other tested combos

In the grand scheme of tested hardware combinations for Thief, this pairing ranks 823rd out of 1,642 combos. That places it almost exactly at the median, which is a noteworthy result given the individual components' percentiles. The CPU sits at the 76th percentile, the GPU at the 82nd percentile, yet the combination lands in the 50th percentile of game-specific results. This gap suggests that Thief's performance is not purely dictated by raw component strength; the game's engine has its own quirks that can suppress or amplify certain hardware pairings.

The nearest rival CPUs listed are all within 0.3% of the 3600's average score, so swapping between those parts would produce negligible changes in synthetic benchmarks. However, the combo rank of 823 out of 1,642 implies that many less powerful GPUs paired with similarly strong CPUs achieve better or comparable FPS in Thief. This could be due to driver overhead, memory latency, or specific engine paths that favor certain architectures. What matters for a builder is that this specific combination is neither a standout nor a laggard; it is a middle-of-the-pack performer in this specific title, despite the GPU's high percentile ranking.

Resolution Scaling — how FPS drops from 1080p to 4K and what it says about the limiting component

The measured FPS data shows a clear scaling trend as resolution increases. Going from 1920x1080 to 2560x1440 at High settings, the average FPS drops from 104.0 to 89.9, a reduction of about 14%. Moving further to 3840x2160 at High, the average drops to 57.9, which is a 44% decrease from the 1080p result. This steep falloff at 4K is a classic sign that the GPU is the limiting component at higher resolutions. The CPU's workload does not change much with resolution, so the FPS decline is almost entirely attributable to the GPU's pixel-pushing capability.

At Low settings, the pattern is similar but less dramatic. The 1080p Low result is 128.3 FPS, dropping to 111.2 at 1440p and 95.3 at 4K. The 4K Low figure is still above 60 FPS, which indicates the GPU has enough raw throughput to handle the game at reduced settings even at the highest resolution. The gap between 1080p and 4K at Low is about 25%, while at High it is 44%. This widening gap with higher settings confirms that pixel fill rate and memory bandwidth (644.6 GB/s on the RX 9070 XT) are the primary constraints at 4K, not the CPU.

Measured FPS Breakdown — resolution by resolution, settings by settings, exact numbers

At 1920x1080, the combo delivers 128.3 FPS on Low, 111.6 on Medium, 104.0 on High, and 101.9 on Ultra. The spread from Low to Ultra is only about 26 FPS, which is a modest penalty for the highest quality settings. This suggests that at 1080p, the game is not heavily taxing the GPU, and the CPU's 12 threads are sufficient to keep frame rates high across all presets. The Ultra preset at 101.9 FPS is still well above the 60 FPS threshold, making 1080p Ultra a comfortable experience.

Moving to 2560x1440, the numbers are 111.2 FPS on Low, 100.9 on Medium, 89.9 on High, and 86.9 on Ultra. The relative drop from 1080p is noticeable but not punishing; even at Ultra, the average stays above 86 FPS. This resolution is the sweet spot for this pairing, balancing visual fidelity with smooth performance. At 3840x2160, the results are 95.3 FPS on Low, 80.5 on Medium, 57.9 on High, and 57.5 on Ultra. The High and Ultra presets both hover near 58 FPS, which is just shy of a locked 60 FPS experience. Low and Medium presets at 4K remain playable, but the Ultra preset's 57.5 FPS indicates the GPU is nearing its limit in this title at maximum settings.

GPU Role — VRAM, clocks, and how they relate to this game's results

The Radeon RX 9070 XT brings 16 GB of GDDR6 memory on a 256-bit bus, with a bandwidth of 644.6 GB/s. For Thief, a game from 2014, VRAM capacity is not a concern; even at 4K Ultra, the game likely uses far less than 16 GB. The memory clock runs at 2518 MHz (20.1 Gbps effective), which provides plenty of throughput for the game's textures. The GPU's boost clock of 2970 MHz and game clock of 2400 MHz are high, and the card's 128 ROPs and 256 TMUs handle the game's draw calls efficiently.

The benchmark scores reflect the GPU's strength: a Passmark G3D score of 26,795 and a 3DMark Steel Nomad DX12 score of 7,260 put it in the 82nd percentile. However, the nearest rivals listed are older or lower-tier parts like the AMD Radeon Pro 575 (avg score 39,703, delta -0.1%) and NVIDIA RTX A500 Mobile (avg score 39,568, delta +0.2%). The RX 9070 XT's average benchmark score of 39,647 is essentially tied with these, which is surprising given the generational difference. The data suggests that in synthetic tests, the RX 9070 XT does not blow away older workstation cards, but in actual Thief gameplay, the 16 GB VRAM and modern architecture ensure no memory-related stutters or texture pop-in.

Settings Recommendations — which preset gives the best experience per the measured rows

For a smooth experience, the data points to 2560x1440 with the High preset as the best balance. At 89.9 FPS, it is well above 60 FPS and only 3 FPS slower than the Ultra preset at that resolution. The jump from High to Ultra at 1440p costs only 3 FPS (86.9 vs 89.9), so if you want the extra visual polish, Ultra is nearly free. However, at 1080p, the Ultra preset at 101.9 FPS is a better value than High at 104.0 FPS, since the difference is negligible and Ultra offers better visuals.

At 4K, the situation is more nuanced. The High preset delivers 57.9 FPS, which is just below the 60 FPS target; the Ultra preset is essentially identical at 57.5 FPS. If you are using a 4K display, the Medium preset at 80.5 FPS is the practical choice, as it provides a comfortable margin over 60 FPS while retaining decent visual quality. The Low preset at 4K hits 95.3 FPS, but the visual downgrade is likely too severe for a stealth game where lighting and shadows matter. The recommendation is to prioritize 1440p High/Ultra for the best experience, or drop to 4K Medium if you must have native 4K resolution. The CPU will not be a limiting factor in any of these scenarios, so the choice comes down entirely to your display's native resolution and your tolerance for sub-60 FPS dips at 4K.