Control

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

The AMD Ryzen 7 5800X is an 8-core, 16-thread desktop processor built on the Zen 3 architecture (codenamed "Vermeer") using TSMC's 7 nm process. It carries a base clock of 3.80 GHz and a boost clock of 4.70 GHz, with a 105 W TDP and support for dual-channel DDR4 memory at a theoretical bandwidth of 51.2 GB/s. The CPU's cache hierarchy includes 64 KB of L1 per core, 512 KB of L2 per core, and a 32 MB shared L3 cache. It also supports ECC memory, uses the AMD Socket AM4, and offers PCIe Gen 4 with 20 lanes (CPU only). The processor was released in November 2020 with a launch MSRP of $449 and remains in active production.

In synthetic threaded workloads, the 5800X posts a 3DMark score of 7,586 in the 16-thread test and 7,578 in the max-threads test, showing near-linear scaling from 8 to 16 threads. Its single-thread 3DMark score is 943, while the 2-thread and 4-thread scores are 1,846 and 3,543 respectively, and the 8-thread score reaches 6,035. Cinebench results reinforce this: R15 multi-core hits 2,608.5, single-core 265.5, while R23 multi-core reaches 15,476 and single-core 1,574.5. Geekbench scores are 11,036 multi-core and 2,032 single-core. PassMark results show a multithread score of 27,732, single-thread of 3,445, and a physics score of 1,364 — the latter being particularly relevant for game physics calculations.

The CPU's percentile vs all CPUs is 81, placing it in the top fifth of all processors tested. Its average benchmark score is 29,123. Nearest rivals include the Intel Core Ultra 5 135H (average score 29,093, delta +0.1%), the Intel Xeon Phi 7210 (29,245, delta -0.4%), the Intel Core i5-13500HX (29,270, delta -0.5%), and the AMD Ryzen 5 5600XT (28,940, delta +0.6%). These deltas are all within 1% of each other, meaning the 5800X sits in a tightly packed performance band where no single rival holds a decisive edge.

For a game like Control, which is an action-adventure title from 2019, the CPU's 16 threads matter because the game's physics and object interactions can spawn multiple concurrent tasks. The 5800X's 8 physical cores with SMT provide enough parallel capacity to avoid becoming the bottleneck at lower resolutions, where frame rates are higher and per-frame CPU work becomes more visible. The near-linear 3DMark scaling from 8 to 16 threads (6,035 to 7,586) indicates the CPU can leverage additional threads effectively, which is useful for background processes while gaming. The single-thread score of 943 in 3DMark and 1,574.5 in Cinebench R23 suggests strong per-core performance, which is critical for games that rely on a main render thread. The L3 cache of 32 MB is generous for a desktop chip and helps with data locality in game worlds.

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

The NVIDIA GeForce RTX 5070 is built on the Blackwell 2.0 architecture with the GB205 chip, manufactured on TSMC's 5 nm process with 31,100 million transistors on a 263 mm² die. It has 6,144 shading units, 192 texture mapping units, 80 ROPs, 48 ray tracing cores, and 192 tensor cores. The base clock is 2,325 MHz with a boost clock of 2,512 MHz, and memory runs at 1,750 MHz (28 Gbps effective). The GPU comes with 12 GB of GDDR7 memory on a 192-bit bus, yielding a bandwidth of 672.0 GB/s. Pixel rate is 201.0 GPixel/s, texture rate is 482.3 GTexel/s, and FP32/FP16 performance is 30.87 TFLOPS (1:1). The card has a 250 W TDP, is dual-slot, uses a single 16-pin power connector, and recommends a 600 W PSU. It interfaces via PCIe 5.0 x16 and outputs include 1x HDMI 2.1b and 3x DisplayPort 2.1b. API support covers DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4. Released in March 2025 with a launch MSRP of 549 USD, it succeeds the GeForce 40 series and precedes the GeForce 60 series.

In synthetic benchmarks, the RTX 5070 scores 5,077 in 3DMark Steel Nomad DX12, 172,660 in Geekbench OpenCL, and 178,923 in Geekbench Vulkan. PassMark results show a G3D score of 29,137, a GPU compute score of 15,787, and DirectX-specific scores of 180 (DX10), 277 (DX11), 108 (DX12), and 320 (DX9). The GPU's percentile vs all GPUs is 82, and its average benchmark score is 40,377. Nearest rivals include the AMD Radeon Pro 580 (average score 40,318, delta +0.1%), AMD Radeon Pro WX 7100 (40,063, delta +0.8%), AMD Radeon Pro 5300 (40,870, delta -1.2%), and NVIDIA RTX A500 Mobile (39,568, delta +2%). Again, all rivals are within 2% of the RTX 5070's average score, indicating that this GPU's raw compute performance is comparable to a set of professional-grade cards.

For Control, the GPU's 12 GB VRAM is more than sufficient for a 2019 title, even at 4K with high textures. The 672.0 GB/s memory bandwidth ensures that high-resolution texture streaming does not become a limiting factor. The boost clock of 2,512 MHz, combined with 80 ROPs, drives the pixel throughput needed for the game's particle effects and environmental destruction. The 48 RT cores are relevant if ray tracing is enabled, though the measured FPS data in this pack does not include ray tracing settings. The 30.87 TFLOPS FP32 figure indicates strong compute headroom for the game's physics and post-processing shaders.

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

Looking at the measured FPS data for Control with this CPU/GPU combo, the scaling across resolutions reveals where the bottleneck lies. At 1080p with Low settings, the average FPS is 206; at 1440p Low, it drops to 156; and at 4K Low, it falls to 105. That is a 49% reduction from 1080p to 4K at Low settings. For High settings, the figures are 130 FPS at 1080p, 98 FPS at 1440p, and 66 FPS at 4K — a 49% drop from 1080p to 4K as well. Medium settings show 146 FPS at 1080p, 111 FPS at 1440p, and 75 FPS at 4K, which is a 49% decline. Ultra settings yield 113 FPS at 1080p, 85 FPS at 1440p, and 58 FPS at 4K, also a 49% reduction.

The consistency of the percentage drop — right around half — across all four settings strongly suggests that the GPU is the primary limiting factor when moving to higher resolutions. The pixel count quadruples from 1080p to 4K, and the frame rate halves, which is exactly what you would expect if the GPU's fill rate and shading throughput are saturated. The CPU's role becomes less important at 4K because the frame rate is lower, giving the 5800X more time per frame to prepare draw calls. At 1080p Low, the 206 FPS average indicates that the CPU can feed the GPU fast enough, but even then the GPU is not fully idle — the 49% drop to 4K is purely GPU-bound. The data implies that for this combo, resolution scaling is nearly linear with pixel count, so the RTX 5070's 12 GB VRAM and 672 GB/s bandwidth are not the limiting factors; instead, it is the raw shading and ROP throughput.

How This Combo Ranks — use comboRankInGame vs other tested combos

In the benchmark database for Control, this specific combination of the AMD Ryzen 7 5800X and NVIDIA GeForce RTX 5070 achieves a rank of 780 out of 3,720 tested combos. That places it in roughly the 79th percentile of all CPU/GPU pairings tested for this game. The rank of 780 is notable because it sits below the top-tier combos (which would include newer or higher-end parts), but it is still well above the median. The combo's percentile vs all CPUs (81) and vs all GPUs (82) are both high, yet the in-game rank of 780/3720 (about 21% from the top) suggests that Control is not heavily CPU-bound; otherwise, the CPU's 81st percentile would push the combo higher. The gap between the component percentiles and the combo rank implies that the GPU's 82nd percentile is the limiting factor, and the CPU's strength is not fully utilized in this title.

Compared to the nearest rivals on the CPU side — such as the Intel Core Ultra 5 135H or the Intel Core i5-13500HX, both with average scores within 0.5% of the 5800X — the combo's rank would likely be similar if paired with the same GPU. On the GPU side, the RTX 5070's rivals like the AMD Radeon Pro 580 (0.1% delta) or the RTX A500 Mobile (2% delta) would produce slightly different in-game results, but the rank would remain in the same ballpark. The fact that this combo ranks 780th out of 3,720 means there are 2,940 combos that perform better, but also 2,940 that perform worse — it is a solidly upper-mid-range pairing for Control.

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

Given the measured FPS data, the best settings for a smooth experience depend on the target resolution and refresh rate. At 1080p, all four presets deliver above 113 FPS average. The Ultra preset produces 113 FPS, which is high enough for a 120 Hz display, but the Medium preset at 146 FPS provides a 29% higher frame rate with only a moderate visual drop. For 1440p, the High preset averages 98 FPS, which is suitable for a 100 Hz monitor, while Medium at 111 FPS is better for a 120 Hz panel. The Ultra preset at 1440p dips to 85 FPS, which is still playable but below the 90 FPS threshold that many gamers prefer. At 4K, the High preset averages 66 FPS, which is borderline for a 60 Hz display; the Medium preset at 75 FPS is a safer choice for maintaining a 60 FPS minimum, as the measured minimum for Medium at 4K is 63 FPS. The Low preset at 4K (105 FPS) is unnecessary unless the user has a 100 Hz+ 4K monitor, but at that point the visual quality loss is significant.

The sweet spot for most users would be Medium at 1440p (111 FPS average, 94 FPS minimum) or High at 1080p (130 FPS average). For 4K, Medium is the recommended preset because it is the only one with a reported minimum FPS (63) that stays above 60, whereas High and Ultra at 4K have no reported minimums but likely dip lower. The data shows that the step from Medium to Ultra costs roughly 22-27% of frame rate at every resolution (e.g., 146 to 113 at 1080p, 111 to 85 at 1440p, 75 to 58 at 4K), so the visual quality gain from Ultra is not worth the performance hit unless the user has a high-refresh 1080p panel.

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

The measured FPS data for Control with the Ryzen 7 5800X and RTX 5070 is as follows:

1080p (1920x1080):

• Low: 206 FPS average (no min/max reported)
• Medium: 146 FPS average, 124 FPS minimum, 168 FPS maximum
• High: 130 FPS average (no min/max reported)
• Ultra: 113 FPS average (no min/max reported)

1440p (2560x1440):

• Low: 156 FPS average (no min/max reported)
• Medium: 111 FPS average, 94 FPS minimum, 127 FPS maximum
• High: 98 FPS average (no min/max reported)
• Ultra: 85 FPS average (no min/max reported)

4K (3840x2160):

• Low: 105 FPS average (no min/max reported)
• Medium: 75 FPS average, 63 FPS minimum, 86 FPS maximum
• High: 66 FPS average (no min/max reported)
• Ultra: 58 FPS average (no min/max reported)

The Medium preset is the only one for which minimum and maximum FPS values are reported at every resolution. At 1080p, the Medium preset's minimum of 124 FPS ensures no drops below 120 FPS, which is ideal for 120 Hz gaming. At 1440p Medium, the minimum of 94 FPS keeps the experience above 90 FPS. At 4K Medium, the minimum of 63 FPS stays above 60, but just barely — the 86 FPS maximum indicates some headroom, but the 12 FPS spread between minimum and average suggests occasional stutter. The Low preset at 1080p (206 FPS) is the only setting that exceeds 200 FPS, making it suitable for 240 Hz monitors, but the visual trade-off is severe. The Ultra preset at 4K (58 FPS average) is the only setting that falls below 60 FPS, so it is not recommended for smooth 4K play. Overall, the Medium preset offers the best balance of performance and stability across all resolutions, as evidenced by the fact that it is the only preset with reported min/max values that never drop below 60 FPS at any resolution.

FAQ

*Q: Is the AMD Ryzen 7 5800X powerful enough for Control at 4K?*

A: Yes. The data shows that with this CPU and the RTX 5070, the combo achieves 66 FPS at 4K High and 75 FPS at 4K Medium. Since the 5800X's 8 cores and 16 threads provide a 3DMark 16-thread score of 7,586, the CPU is not the bottleneck at 4K — the GPU's 82nd percentile ranking is what limits frame rates.

*Q: How does the RTX 5070's VRAM affect Control at high resolutions?*

A: The GPU has 12 GB of GDDR7 memory on a 192-bit bus with 672.0 GB/s bandwidth. For a 2019 game like Control, 12 GB is ample even at 4K Ultra, as evidenced by the fact that the 4K Ultra average FPS (58) is limited by shading throughput, not memory capacity. The memory bandwidth is more than sufficient for the game's textures.

Q: What is the best settings preset for a 60 Hz 4K monitor?

A: The Medium preset at 4K is the safest choice, as it yields an average of 75 FPS with a minimum of 63 FPS. The High preset at 4K (66 FPS average) has no reported minimum, so it may dip below 60. Ultra at 4K (58 FPS) is below 60 FPS, so it should be avoided for a 60 Hz display.

*Q: How does this combo rank among all tested CPU/GPU pairings for Control?*

A: The combination ranks 780th out of 3,720 tested combos, placing it in the top 21% of all pairings. This is consistent with the CPU's 81st percentile and GPU's 82nd percentile, though the in-game rank is slightly lower, suggesting the game is more GPU-limited than the component percentiles would imply.

*Q: Does the CPU's single-thread performance matter for Control?*

A: Yes, but only at lower resolutions. The 5800X's Cinebench R23 single-core score is 1,574.5, and its 3DMark single-thread score is 943. At 1080p Low, the combo reaches 206 FPS, which is high enough that the main render thread's performance becomes relevant. At 4K, the frame rate drops to 105 FPS at Low, so the CPU has more time per frame and single-thread performance matters less.

Q: Is the RTX 5070's performance comparable to its nearest rivals?

A: The GPU's average benchmark score is 40,377, while its nearest rival, the AMD Radeon Pro 580, scores 40,318 (delta +0.1%). The AMD Radeon Pro WX 7100 scores 40,063 (+0.8%), the AMD Radeon Pro 5300 scores 40,870 (-1.2%), and the NVIDIA RTX A500 Mobile scores 39,568 (+2%). All are within 2% of the RTX 5070, so the performance gap is negligible in synthetic tests, and in-game results would be similar.