Control

Control’s benchmark results for the Intel Core i7-12700K and NVIDIA GeForce RTX 5090 D combination show a system that scales predictably with resolution, though the pattern reveals where the performance ceiling sits. At 1920x1080 with Low settings, the combo delivers 268 FPS average, which drops to 188 FPS at High settings and 163 FPS at Ultra. Moving to 2560x1440, the same settings tier produces 225 FPS at Low, 142 FPS at High, and 124 FPS at Ultra. At 3840x2160, the numbers fall further: 152 FPS at Low, 96 FPS at High, and 83 FPS at Ultra. The gap between 1080p and 4K at High settings is 92 FPS, representing a 49% reduction in frame rate. This scaling curve is steep, but the absolute values remain high enough to suggest that neither component is being pushed to its absolute limit at lower resolutions. The 1080p to 1440p drop at High settings is 46 FPS, while the 1440p to 4K drop is another 46 FPS, which indicates a consistent, linear degradation as pixel count increases. That linearity points to the GPU becoming the primary constraint at higher resolutions, as the CPU’s workload does not scale with pixel count in the same way. The RTX 5090 D’s 32 GB of GDDR7 memory on a 512-bit bus with 1.79 TB/s bandwidth is more than sufficient for Control’s textures and effects, so the frame rate declines are not due to memory capacity or bandwidth starvation. Instead, the data shows a system where the GPU’s raw compute—104.8 TFLOPS FP32—is the limiting factor at 4K, while at 1080p the CPU’s ability to feed frames becomes more relevant.

Resolution Scaling

The FPS progression across resolutions tells a clear story about where the bottleneck shifts. At 1920x1080, the difference between Low and Ultra settings is 105 FPS (268 vs 163), which is a 39% performance hit from the highest quality preset. The same comparison at 2560x1440 shows a 101 FPS gap (225 vs 124), nearly identical in absolute terms. At 3840x2160, the Low to Ultra spread narrows to 69 FPS (152 vs 83), a 45% reduction. This narrowing at 4K is the signature of a GPU-bound scenario—when the render resolution increases, the CPU has more time to prepare frames, so the relative cost of higher settings becomes less about CPU overhead and more about GPU fill rate and shading complexity. The RTX 5090 D’s 176 ROPs and 1,636.8 GTexel/s texture rate handle the increased pixel workload, but the 423.6 GPixel/s pixel rate is finite, and Control’s particle effects and physics-based destruction at Ultra settings push against that limit.

The 1080p Medium result of 211 FPS average, with a minimum of 180 FPS and maximum of 243 FPS, shows a frame time consistency that suggests the CPU is not causing hitches. The i7-12700K’s 12 cores and 20 threads, with a boost clock of 5.00 GHz, provide ample headroom for Control’s single-threaded game logic and multi-threaded physics. At 1440p Medium, the average drops to 160 FPS with a minimum of 136 FPS and maximum of 184 FPS, and at 4K Medium the numbers are 108 FPS average, 92 FPS minimum, and 124 FPS maximum. The minimum FPS values at Medium settings—180, 136, and 92—scale almost perfectly with the average FPS, which indicates that the system is not experiencing sudden frame time spikes. This is a well-balanced pairing where the GPU becomes the sole limiter as resolution climbs, but the CPU never falls behind enough to create stutter.

The 4K High result of 96 FPS average is particularly telling. It sits between the Medium 108 FPS and Ultra 83 FPS, showing that Control’s High preset is a reasonable middle ground for this hardware. The 4K Low result of 152 FPS demonstrates that even at the most demanding resolution, the GPU can push past 144 Hz when settings are relaxed. For a 4K 120 Hz display, the Medium preset at 108 FPS is the sweet spot, while Ultra at 83 FPS would require VRR to feel smooth. The 1440p results are uniformly high—all presets exceed 120 FPS—making this resolution the practical maximum for high refresh rate gaming with this combo. At 1080p, the Low preset’s 268 FPS and Medium’s 211 FPS are beyond the refresh rate of most monitors, meaning the CPU and GPU are both underutilized and the system is effectively waiting on the display.

How This Combo Ranks

In the benchmark database, this specific CPU and GPU pairing ranks 26th out of 2888 tested combos for Control. That places it in the top 0.9% of all configurations, which is an elite tier given the sheer number of possible combinations. The rank reflects the raw performance ceiling of the RTX 5090 D and the i7-12700K’s ability to keep up, but it does not guarantee that every game will see the same relative position. For Control, the combination is clearly near the top of the heap, and the measured FPS data corroborates that ranking—1080p Ultra at 163 FPS and 4K High at 96 FPS are numbers that only a handful of GPUs can achieve. The 92nd percentile for the GPU alone, compared to all GPUs, and the 84th percentile for the CPU, compared to all CPUs, suggest that the GPU is the stronger component in this pairing. That asymmetry is reflected in the ranking: the combo outperforms configurations with weaker GPUs but similar CPUs, while it would not surpass systems with the same GPU but a faster CPU in CPU-bound scenarios.

The nearest rivals for the GPU, based on average benchmark scores, include the AMD Radeon RX 6650M XT with a delta of 1.1% and the AMD Radeon RX 6850M XT with a delta of -1.6%. These are mobile GPUs, which is unusual, but the deltaPct values indicate that the RTX 5090 D’s average score of 77712 is very close to those parts in synthetic tests. In Control specifically, the measured FPS suggests the desktop RTX 5090 D pulls ahead due to its 32 GB memory and higher clocks, but the synthetic benchmark proximity shows that raw compute is not the only factor. The CPU’s nearest rivals—the Intel Core i5-13600T with a delta of -0.1%, the i7-12700KF with -0.2%, the i7-13700T with -0.3%, and the Intel Core 5 213PE with -0.4%—are all within a fraction of a percent in average score. This means the i7-12700K is not leaving performance on the table; its 9979 max thread score and 19784 Cinebench R23 multicore score are representative of its class, and the combo rank of 26 is more a function of the GPU’s dominance than the CPU’s unique contribution.

GPU Role

The NVIDIA GeForce RTX 5090 D is the dominant component in this pairing, and its specifications explain the measured FPS outcomes. The GPU has 32 GB of GDDR7 memory on a 512-bit bus, yielding 1.79 TB/s of bandwidth. Control is not a memory-heavy game by modern standards, but this capacity ensures that even at 4K Ultra with high-resolution textures, there is no VRAM pressure. The 92,200 million transistors on a 750 mm² die, built on a 5 nm process at TSMC, provide the computational foundation for the 104.8 TFLOPS FP32 throughput. That number is the key to understanding the 4K results: at 3840x2160, the GPU is rendering 8.3 million pixels per frame, and the shading units (21,760) must process each one. The 2407 MHz boost clock is moderate for a flagship, but the sheer number of cores compensates. The 170 RT cores and 680 tensor cores are relevant for Control’s ray tracing effects, though the measured FPS data does not specify whether ray tracing is enabled—the settings tiers are simply Low, Medium, High, and Ultra.

The GPU’s benchmark scores place it at the 92nd percentile, with a 3DMark Steel Nomad DX12 score of 14326 and a Passmark G3D score of 44065. These are synthetic results, but they correlate with the in-game performance. The 4K Ultra result of 83 FPS average is the most demanding scenario in the dataset, and it shows that the GPU is not able to maintain 100+ FPS at the highest preset. The 4K High result of 96 FPS is closer, and the 4K Medium result of 108 FPS shows that reducing settings from Ultra to Medium yields a 30% FPS increase. The GPU’s 176 ROPs are the final stage of the render pipeline, and at 4K, the pixel rate of 423.6 GPixel/s becomes a hard limit. For comparison, the 1080p Medium result of 211 FPS means the GPU is producing 211 frames per second, each with 2.1 million pixels, for a total of 443 million pixels per second—just above the pixel rate limit, which is why the GPU is fully utilized at that setting. At 4K Medium, 108 FPS at 8.3 million pixels per frame equals 896 million pixels per second, which is more than double the pixel rate, indicating that the GPU is not purely pixel-bound at that resolution and that shading and texture work are also contributing to the frame time.

Measured FPS Breakdown

The complete measured FPS dataset provides a granular view of performance. At 1920x1080, the Low preset delivers 268 FPS average, the Medium preset delivers 211 FPS with a minimum of 180 FPS and maximum of 243 FPS, the High preset delivers 188 FPS, and the Ultra preset delivers 163 FPS. The spread from Low to Ultra is 105 FPS, and the Medium settings show a 63 FPS range between minimum and maximum, which is a 30% variance—this is normal for a scene with variable complexity. At 2560x1440, the numbers are 225 FPS at Low, 160 FPS at Medium with a minimum of 136 FPS and maximum of 184 FPS, 142 FPS at High, and 124 FPS at Ultra. The Medium variance is 48 FPS, or 30%, matching the 1080p pattern. The step from Medium to High at 1440p is 18 FPS, and from High to Ultra is another 18 FPS, showing a consistent cost for each quality tier. At 3840x2160, the Low preset delivers 152 FPS, Medium delivers 108 FPS with a minimum of 92 FPS and maximum of 124 FPS, High delivers 96 FPS, and Ultra delivers 83 FPS. The Medium variance at 4K is 32 FPS, or 30%, which is the same relative spread as the lower resolutions, indicating that the GPU’s frame pacing does not degrade with resolution.

The Medium settings are the only tier with minimum and maximum FPS recorded, and they provide insight into frame time stability. At 1080p, the 180 FPS minimum means the system never drops below 180 FPS, which is a 33% margin above the 144 Hz threshold. At 1440p, the 136 FPS minimum is just below 144 Hz, meaning a 144 Hz display would see occasional dips. At 4K, the 92 FPS minimum is comfortably above 60 Hz but would require a 120 Hz display with VRR to avoid tearing. The High and Ultra presets do not have minimum FPS data, but the averages suggest that 4K Ultra at 83 FPS is the most demanding scenario, and it is unlikely that the minimum stays above 60 FPS given the 30% variance observed in the Medium preset. The 1080p Low result of 268 FPS is the highest in the dataset, and it indicates that the CPU is capable of feeding the GPU at very high frame rates—the i7-12700K’s 5.00 GHz boost clock and 2065 3DMark 2-thread score are sufficient to avoid a CPU bottleneck at this resolution.

FAQ

Q: What is the highest average FPS this combo achieves in Control?

A: The highest average FPS is 268, achieved at 1920x1080 with Low settings.

Q: How does the combo perform at 4K with Ultra settings?

A: At 3840x2160 with Ultra settings, the average FPS is 83, which is the lowest result in the measured dataset.

Q: Is the RTX 5090 D’s 32 GB VRAM a limiting factor in Control?

A: No, the 32 GB of GDDR7 memory on a 512-bit bus is far more than Control requires, and the FPS drops at 4K are due to compute limits, not memory capacity.

Q: What is the frame time consistency at Medium settings?

A: At Medium settings, the minimum and maximum FPS show a consistent 30% variance across all resolutions: 180-243 FPS at 1080p, 136-184 FPS at 1440p, and 92-124 FPS at 4K.

Q: How does the combo rank compared to other tested configurations?

A: It ranks 26th out of 2888 tested combos for Control, placing it in the top 0.9% of all configurations.

Q: Does the CPU bottleneck the GPU at any resolution?

A: At 1080p Low, the 268 FPS average is so high that the CPU’s single-thread performance becomes the limiting factor, but at 1440p and above, the GPU is the primary constraint.

CPU Role

The Intel Core i7-12700K is a 12-core, 20-thread processor based on the Alder Lake architecture, built on Intel’s 10 nm process. It has a base clock of 3.60 GHz and a boost clock of 5.00 GHz, with 25 MB of shared L3 cache. The CPU’s benchmark scores show a strong single-thread performance—1850.5 in Cinebench R23 single-core and 2308 in Geekbench single-core—which is critical for Control’s game logic and draw call processing. The multi-thread scores are also robust, with 19784 in Cinebench R23 multicore and 34404 in Passmark multithread, which helps with the game’s physics and particle simulations. The 3DMark 16-thread score of 9276 and max-thread score of 9979 indicate that the CPU scales well with additional threads, though Control does not heavily utilize more than a handful of cores for its core loop. The 84th percentile ranking among all CPUs reflects that this is a solid performer, but not an outlier—the nearest rivals are all within 0.4% of its average score, meaning the 12700K is not a unique enabler of high FPS in Control.

The CPU’s role in the measured FPS is most evident at 1080p Low, where the 268 FPS average approaches the limits of what the GPU can render when settings are minimal. At this resolution, the CPU must prepare 268 frames per second, which requires a certain amount of single-threaded throughput. The 5.00 GHz boost clock provides that, but the data shows that the GPU is still the bottleneck even at Low settings—the 1080p Low result is 76 FPS higher than the 1440p Low result, which is a 34% drop that cannot be explained by CPU workload alone. The CPU’s 20 threads and 12 cores are more than sufficient for Control, and the 25 MB L3 cache helps with the game’s asset streaming. The 3DMark 2-thread score of 2065 suggests that the CPU’s two fastest cores are capable of handling the game’s primary thread, while the additional cores handle background tasks and physics. In summary, the i7-12700K provides enough headroom to avoid becoming a bottleneck at any resolution tested, but it does not contribute to the FPS gains at higher resolutions—those are entirely dependent on the RTX 5090 D’s compute power. The combo rank of 26 out of 2888 confirms the GPU’s dominance rather than the CPU’s exceptional performance, and for Control specifically, the CPU is a capable partner that stays out of the way.