Control

# Control — Intel Core i7-12700K + NVIDIA GeForce RTX 4070

The combination of Intel's 12-core Core i7-12700K and NVIDIA's GeForce RTX 4070 delivers a robust Control experience across all tested resolutions, with the data showing a clear progression from 193 FPS at 1080p Low to 54 FPS at 4K Ultra. This pairing ranks 904th out of 3,720 tested CPU-GPU combinations in this game, placing it comfortably above the median for the full dataset. The benchmark results indicate that the system handles Control's physics-heavy combat and destructible environments with headroom at lower settings, while 4K Ultra pushes the GPU to its limits.

Resolution Scaling

The measured FPS data reveals how Control scales across the three tested resolutions. At 1080p Low, the system produces 193 FPS, which drops to 146 FPS at 1440p Low — a 24% reduction — and then to 98 FPS at 4K Low, a further 33% decline. This pattern of diminishing returns as resolution increases is typical of a GPU-bound scenario at lower quality presets.

Moving to High settings, the scaling is more gradual: 122 FPS at 1080p, 92 FPS at 1440p (25% drop), and 62 FPS at 4K (33% drop from 1440p). The consistency of these percentage drops across Low and High presets suggests the RTX 4070's rendering pipeline scales predictably with pixel count, with no sudden bottleneck emerging at any single resolution step.

The Medium preset shows similar behavior, with 137 FPS at 1080p falling to 104 FPS at 1440p (24% drop) and then to 70 FPS at 4K (33% drop). Ultra settings follow the same trajectory: 106 FPS at 1080p, 80 FPS at 1440p (25% drop), and 54 FPS at 4K (33% drop). The uniformity of these scaling ratios — roughly 75% retention from 1080p to 1440p and 67% retention from 1440p to 4K — indicates that the limiting component remains consistent across all settings.

What this scaling tells us about the system's balance: at 1080p, the frame rates are high enough (106–193 FPS) that the CPU's 12 cores and 20 threads are likely keeping pace without constraint, given the processor's 3DMark single-thread score of 1,043 and multi-thread score of 9,979. At 4K, the RTX 4070's 12 GB of GDDR6X memory and 504.2 GB/s bandwidth become the dominant factor, as evidenced by the frame rates clustering between 54 and 98 FPS depending on preset.

Notably, the data shows no instance where a resolution upgrade causes a catastrophic frame rate collapse. The worst-case scenario — 4K Ultra at 54 FPS — is still playable, while 4K High at 62 FPS approaches the 60 FPS threshold many players target. The 1440p results, ranging from 80 to 146 FPS, offer the most flexible operating range for high-refresh-rate displays.

GPU Role

The RTX 4070's specifications directly inform its Control performance. With 5,888 shading units, 184 texture mapping units, and 64 ROPs, the GPU processes the game's particle effects and physics-based debris efficiently. The boost clock of 2,475 MHz (base 1,920 MHz) provides the raw throughput needed for Control's frequent slow-motion sequences and telekinetic object manipulation.

Memory configuration plays a critical role at higher resolutions. The 12 GB GDDR6X frame buffer, operating at 1,313 MHz (21 Gbps effective) across a 192-bit bus, delivers 504.2 GB/s of bandwidth. This proves sufficient for 4K Ultra (54 FPS) without any measured frame drops below the average, though the margin is thin. The Medium preset at 4K shows a minimum of 59 FPS against a 70 FPS average, indicating some frame-time variance but no severe stuttering.

Ray tracing is where this GPU's Ada Lovelace architecture could shine, given the 46 RT cores and 184 tensor cores present. However, the measured data does not include ray-traced settings, so we can only infer from the standard rasterization results. The 29.15 TFLOPS FP32 compute rating suggests ample headroom for compute-heavy effects, which Control uses extensively for its procedural destruction and particle simulations.

The GPU's benchmark scores align with its gaming performance: a PassMark G3D score of 26,927 and a 3DMark Steel Nomad DX12 score of 3,854 place it in the 81st percentile of all GPUs. Its nearest rivals — the NVIDIA Tesla P4 (0.1% ahead), AMD Radeon RX Vega 56 (0.4% behind), and NVIDIA GeForce RTX 4080 Mobile (1.3% behind) — all cluster within a narrow performance band, confirming that the RTX 4070's positioning in Control's results is consistent with its broader benchmark profile.

The 200 W TDP and dual-slot cooler design suggest sustained performance without thermal throttling concerns during extended play sessions, though the data does not include temperature measurements. The 1x 16-pin power connector and 550 W suggested PSU indicate modest system power requirements for this class of GPU.

Settings Recommendations

Based on the measured FPS rows, the optimal preset depends on the target display. For 1080p displays, every preset exceeds 100 FPS — even Ultra delivers 106 FPS — making Ultra the clear choice for maximum visual fidelity without sacrificing frame rate. The 137 FPS Medium average with a 116 FPS minimum provides a comfortable buffer for competitive play, though the 193 FPS Low preset will fully utilize 144 Hz displays.

At 1440p, High settings (92 FPS) offer the best balance of visual quality and performance for 60 Hz displays, while Medium (104 FPS) suits 100 Hz monitors. Ultra at 1440p (80 FPS) remains viable for 75 Hz panels, but the 12 FPS difference between Ultra and High is significant enough that most users would prefer High for smoother motion. Low settings at 1440p (146 FPS) are only recommended for esports-style play where frame rate trumps all else.

For 4K, the choices become more constrained. High settings (62 FPS) barely clears 60 FPS and represents the sweet spot for quality-conscious players. Medium (70 FPS) provides a more comfortable margin above 60 FPS while still looking substantially better than Low (98 FPS). Ultra at 4K (54 FPS) falls below the 60 FPS threshold and will show visible judder on standard 60 Hz displays; it is only recommended for players who prioritize image quality over smoothness and can tolerate sub-60 frame rates.

The stand-out recommendation from the data is 1440p High: at 92 FPS, it delivers near-Ultra visuals with a 15% performance margin over Ultra's 80 FPS. Similarly, 1080p Ultra at 106 FPS demonstrates that this CPU-GPU pairing has enough headroom to run maximum settings at full HD without compromise.

Measured FPS Breakdown

1920x1080:

• Low: 193 FPS average — the highest frame rate recorded across all tests, indicating the system is not constrained at this resolution.
• Medium: 137 FPS average (116 minimum, 157 maximum) — a 29% reduction from Low, with a healthy frame-time distribution.
• High: 122 FPS average — only 11% slower than Medium, showing diminishing returns from quality increases.
• Ultra: 106 FPS average — the most demanding 1080p preset still exceeds 100 FPS, making it fully playable on high-refresh displays.

2560x1440:

• Low: 146 FPS average — a 24% drop from 1080p Low, confirming GPU scaling with pixel count.
• Medium: 104 FPS average (88 minimum, 119 maximum) — the minimum of 88 FPS indicates occasional dips but remains smooth.
• High: 92 FPS average — the closest any preset comes to the 60 FPS threshold while being clearly above it.
• Ultra: 80 FPS average — a 25% reduction from High, making this the first preset where frame rate dips below 90.

3840x2160:

• Low: 98 FPS average — surprisingly high for 4K, suggesting the game's Low preset reduces GPU load substantially.
• Medium: 70 FPS average (59 minimum, 80 maximum) — the minimum of 59 FPS shows the GPU nearing its limit.
• High: 62 FPS average — just 2 FPS above the 60 FPS mark, making this the borderline playable preset.
• Ultra: 54 FPS average — the only preset below 60 FPS in the entire dataset, representing the worst-case scenario.

The complete dataset shows a monotonic relationship: every resolution step down reduces FPS, and every quality increase at a fixed resolution reduces FPS. The gaps between presets at each resolution are consistent — roughly 25-30% from Low to Medium, 10-12% from Medium to High, and 13-15% from High to Ultra — indicating well-balanced settings scaling.

FAQ

Q: Can this system run Control at 4K 60 FPS?

A: Yes, but only at Medium settings, which average 70 FPS (59 minimum). High settings average 62 FPS, which is technically above 60 but with no margin. Ultra settings average 54 FPS, which falls below the 60 FPS threshold.

Q: What is the best preset for a 144 Hz 1080p monitor?

A: Medium settings provide 137 FPS average with a 116 FPS minimum, ensuring the display rarely drops below its refresh rate. Low settings (193 FPS) would also work but sacrifice visual quality. Ultra (106 FPS) cannot consistently saturate a 144 Hz panel.

Q: How does the RTX 4070 compare to its nearest GPU rivals in this context?

A: The RTX 4070's average benchmark score of 37,648 is within 0.4% of the AMD Radeon RX Vega 56 (37,507) and 1.3% behind the NVIDIA GeForce RTX 4080 Mobile (38,135). It is essentially tied with the NVIDIA Tesla P4 (37,628), which is 0.1% faster.

Q: Does the CPU bottleneck the GPU in Control?

A: The data suggests no significant CPU bottleneck. At 1080p Low, the system reaches 193 FPS, and at 1080p Ultra it maintains 106 FPS, with the Intel Core i7-12700K's 3DMark multi-thread score of 9,979 and Cinebench R23 multi-core score of 19,784 providing ample headroom for the game's physics engine.

Q: What is the frame rate difference between Low and Ultra at each resolution?

A: At 1080p, the difference is 87 FPS (193 vs 106). At 1440p, the difference is 66 FPS (146 vs 80). At 4K, the difference is 44 FPS (98 vs 54). The shrinking gap at higher resolutions confirms the GPU becomes the limiting factor.

Q: Is the 12 GB VRAM sufficient for Control at maximum settings?

A: Yes. The 4K Ultra preset averages 54 FPS with no reported minimum FPS drops, indicating the 12 GB GDDR6X frame buffer is not being exhausted. The 504.2 GB/s memory bandwidth is adequate for the game's texture streaming at all tested settings.

CPU Role

The Intel Core i7-12700K brings 12 cores (with 20 threads) based on the Alder Lake architecture, running at a 3.60 GHz base clock and 5.00 GHz boost clock. This 10 nm processor's performance profile — a 3DMark single-thread score of 1,043 and multi-thread score of 9,979 — suggests it provides consistent frame delivery in Control's CPU-intensive moments, such as when multiple objects are being telekinetically thrown or when the environment collapses.

The CPU's Cinebench R23 results (19,784 multi-core, 1,850.5 single-core) and Geekbench scores (15,790 multi-core, 2,308 single-core) place it in the 84th percentile of all CPUs, with an average benchmark score of 35,287. Its nearest rivals — the Intel Core i5-13600T (-0.1%), Intel Core i7-12700KF (-0.2%), Intel Core i7-13700T (-0.3%), and Intel Core 5 213PE (-0.4%) — all score within 0.4% of each other, indicating the 12700K sits at a tightly competitive performance tier.

In Control specifically, the CPU's PassMark physics score of 1,790 and integer math score of 114,264 are relevant to the game's real-time physics calculations. The 25 MB shared L3 cache and 1.25 MB per-core L2 cache help maintain smooth frame pacing, as evidenced by the 1080p Low result of 193 FPS where the GPU is not the bottleneck.

The 125 W TDP and Intel Socket 1700 platform support DDR4 and DDR5 memory, though the measured data does not specify which memory type was used. The UHD Graphics 770 integrated GPU is present but irrelevant to gaming performance with the discrete RTX 4070 installed. The CPU's 20 PCIe Gen 4 lanes provide sufficient bandwidth for the GPU's PCIe 4.0 x16 interface.

At lower resolutions and settings, the CPU's role becomes more prominent. The 193 FPS at 1080p Low demonstrates that the 12-core processor can feed frames faster than most displays can show, while the 3DMark 16-thread score of 9,276 indicates strong multi-threaded performance for background tasks and game engine overhead. However, the data shows that resolution increases quickly shift the bottleneck to the GPU, with 4K results consistently 50-70% lower than 1080p results regardless of settings preset.

The processor's single-thread performance (Cinebench R15 score of 272, PassMark single-thread score of 4,013) is adequate for Control's main game thread, which typically handles player input and AI logic. The 20 threads ensure that the game's secondary threads for physics, audio, and streaming never starve for CPU resources, a critical factor in a game known for its complex interactive environments.