Control

Control’s benchmark results with the Intel Core Ultra X7 368H and NVIDIA GeForce RTX 5090 show a system that is heavily GPU-bound at higher resolutions but reveals a significant CPU dependency at 1080p. The data indicates a strong, playable experience across all settings, with the most demanding Ultra preset still delivering an average of 111 FPS at 1920x1080. This pairing lands in the 80.5th percentile of all tested combos for this game, placing it 563rd out of 2888 systems.

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

The Intel Core Ultra X7 368H is a 16-core, 16-thread mobile processor based on the Panther Lake architecture, built on Intel's 3 nm process. It operates at a base clock of 2.00 GHz and a boost clock of 5.00 GHz, with a 25 W TDP that classifies it as an efficiency-focused mobile part. The processor includes 18 MB of shared L3 cache and 2.5 MB of L2 cache per core, with its memory bandwidth capped at 153.6 GB/s over a dual-channel LPDDR5X interface.

In synthetic benchmarks, this CPU achieves a Cinebench R23 multi-core score of 28221 and a single-core score of 3984. Its Passmark multi-thread score is 32956, with a single-thread score of 4005. These results place the chip at the 87th percentile against all CPUs, with an average benchmark score of 40518. Its nearest rivals in aggregate performance are the Intel Core 7 253PE (0.1% lower), Intel Core 5 223PE (0.2% lower), AMD Ryzen 9 7940H (0.2% higher), and Intel Xeon 6507P (0.2% higher) — the deltaPct values show the 368H is essentially tied with all of them, differing by no more than two-tenths of a percent.

In Control, the CPU’s role becomes evident when observing the frame rate scaling from 1440p to 1080p. At 3840x2160 with High settings, the average FPS is 65; dropping to 2560x1440 raises that to 97 FPS; and at 1920x1080, it reaches 128 FPS. The jump from 1440p to 1080p is a 31 FPS increase (from 97 to 128), which is a substantial gain that indicates the processor is capable of feeding the GPU enough data at lower resolutions to prevent a hard bottleneck. The single-core performance, evidenced by the 3984 Cinebench R23 single-core score, is crucial here, as Control’s physics and scripting systems rely heavily on strong per-thread throughput. The 16 threads also ensure that background tasks and the game’s multi-threaded rendering pipeline have ample resources, though the data suggests that beyond 1080p, the GPU becomes the primary limiter, as the FPS gains from lowering resolution diminish.

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

The NVIDIA GeForce RTX 5090 is a flagship mobile GPU based on the Blackwell 2.0 architecture, built on TSMC's 5 nm process with a massive 92,200 million transistors on a 750 mm² die. It features 32 GB of GDDR7 memory on a 512-bit bus, delivering a bandwidth of 1.79 TB/s. The GPU operates at a base clock of 2017 MHz with a boost clock of 2407 MHz, and the memory runs at 1750 MHz (28 Gbps effective). The card has 21760 shading units, 680 TMUs, and 176 ROPs, with 170 RT cores and 680 tensor cores. Its compute capabilities are substantial: 104.8 TFLOPS FP32 and a pixel rate of 423.6 GPixel/s.

In synthetic workloads, the RTX 5090 scores 18355 in 3DMark Steel Nomad DX12, 334370 in Geekbench OpenCL, and 376728 in Geekbench Vulkan. Its Passmark G3D score is 39650, with a GPU compute score of 26756. The GPU holds the 92nd percentile against all GPUs, with an average benchmark score of 79842. Its nearest rivals are the NVIDIA Tesla P100 PCIe 16 GB (0.3% higher), Tesla P100 PCIe 12 GB (0.6% higher), AMD Radeon RX 6850M XT (1.1% higher), and AMD Radeon Pro Vega 64X (1.4% lower). The deltaPct values indicate the RTX 5090 is within a narrow band of these competitors, though the 5090’s 32 GB VRAM is far larger than the 12-16 GB found on the Tesla parts.

In Control, the GPU’s dominance is clear at 4K. At 3840x2160 with Ultra settings, the average FPS is 57, while High settings yield 65 FPS. The step to Medium raises it to 73 FPS, and Low settings hit 103 FPS. The VRAM capacity is more than sufficient for Control’s textures at any resolution, as the game is not particularly memory-hungry by modern standards, but the raw compute power of the 5090 is what drives these numbers. The 32 GB GDDR7 allocation ensures no texture swapping or stutter at 4K Ultra, which is reflected in the consistent averages. The GPU’s 104.8 TFLOPS FP32 performance is the key reason the 4K Low preset can exceed 100 FPS, a figure that would be unattainable with a lesser card. The data suggests that at 4K, the system is entirely GPU-bound, as the CPU’s influence on frame rates is minimal when the resolution is this demanding.

FAQ — 4-6 Q&A pairs answerable from FACT PACK data

Q: What is the highest average frame rate the combo achieves in Control at 1920x1080?

A: The highest average frame rate at 1920x1080 is 203 FPS, achieved with the Low settings preset. The Medium preset yields 144 FPS, High yields 128 FPS, and Ultra yields 111 FPS.

Q: How does the combo rank among all tested systems for this game?

A: The combo ranks 563rd out of 2888 tested combinations, placing it in the top 19.5% of all systems tested for Control.

Q: What is the minimum FPS recorded in the measured data?

A: The lowest minimum FPS recorded is 62 FPS, which occurs at 3840x2160 with Medium settings. At 2560x1440 Medium, the minimum is 93 FPS, and at 1920x1080 Medium, the minimum is 122 FPS.

Q: Is the CPU or GPU the limiting factor at 4K resolution?

A: The data indicates the GPU is the limiting factor at 4K. The average FPS at 3840x2160 ranges from 57 (Ultra) to 103 (Low), and the scaling from 1440p to 4K shows a much larger FPS drop than the scaling from 1080p to 1440p, which is characteristic of GPU-bound performance.

Q: What is the performance delta between the CPU and its closest rival?

A: The Intel Core Ultra X7 368H has an average benchmark score of 40518, which is 0.1% lower than the Intel Core 7 253PE (40557) and 0.2% higher than the AMD Ryzen 9 7940H (40431). The largest delta is 0.6% against the Intel Core 5 223PE (40585).

Q: Does the GPU outperform its nearest rival in synthetic benchmarks?

A: The RTX 5090’s average benchmark score is 79842. It trails the AMD Radeon Pro Vega 64X (80959) by 1.4%, but it leads the AMD Radeon RX 6850M XT (78940) by 1.1%. Its performance is within 0.6% of the NVIDIA Tesla P100 variants.

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

For a 1920x1080 display, the High preset offers the best balance. The High preset delivers an average of 128 FPS, which is well above the 111 FPS of Ultra, while the Medium preset provides 144 FPS and Low hits 203 FPS. The step from High to Ultra costs 17 FPS, which is a 13.3% reduction, and the visual fidelity gains from Ultra are typically marginal in Control’s atmospheric lighting. The Medium preset’s 144 FPS is smoother, but the quality drop from High to Medium is more noticeable in texture detail and draw distances. Therefore, High is the sweet spot for 1080p, ensuring a high refresh rate experience without sacrificing significant visual quality.

At 2560x1440, the recommendation shifts to Medium. The Medium preset averages 109 FPS with a minimum of 93 FPS, which guarantees a smooth experience on 144Hz monitors. The High preset yields 97 FPS, which is still playable but leaves less headroom for frame drops in combat-heavy scenes. The Ultra preset drops to 84 FPS, which is below the 90 FPS threshold many consider ideal for fast-paced action. Low at 153 FPS is uncapped but loses too much visual fidelity. Medium provides the best compromise between the 153 FPS of Low and the 84 FPS of Ultra.

For 3840x2160, the Medium preset is the only viable option for a consistent 60 FPS experience. It averages 73 FPS with a minimum of 62 FPS, which stays above the 60 FPS target. The High preset averages 65 FPS, but its lack of a recorded minimum FPS suggests it may dip below 60 FPS in intense moments. Ultra at 57 FPS is below the 60 FPS threshold and is not recommended for 4K gaming. Low at 103 FPS is smooth but sacrifices significant graphical detail. Medium is the clear choice for 4K, offering the only preset that maintains a playable frame rate while retaining acceptable visual quality.

How This Combo Ranks — use comboRankInGame vs other tested combos

The combination of the Intel Core Ultra X7 368H and NVIDIA GeForce RTX 5090 achieves a rank of 563 out of 2888 tested combos in Control. This places the system in the 80.5th percentile, meaning it outperforms approximately 80.5% of all other tested configurations. The rank is a composite measure that accounts for both the CPU and GPU’s contributions to the game’s performance, and it reflects the system’s ability to maintain high frame rates across resolutions.

Given the GPU’s 92nd percentile ranking against all GPUs and the CPU’s 87th percentile ranking against all CPUs, the combo’s 80.5th percentile game-specific rank is slightly lower than either component’s individual standing. This discrepancy suggests that Control’s engine does not scale perfectly with the raw power of this hardware, or that the mobile CPU’s 25 W TDP limits sustained performance in a way that synthetic benchmarks do not capture. The rank of 563 implies that there are 562 combinations that perform better in this specific title, many of which likely pair desktop-class CPUs with the same or similar GPUs. The deltaPct against the CPU’s nearest rivals (all within 0.2%) shows that the processor choice is not the differentiator here; the GPU’s performance is what drives the rank. However, the fact that the combo does not crack the top 500 suggests that the 368H’s power constraints might be holding back the 5090 in CPU-bound scenarios at lower resolutions, even though the GPU dominates at 4K.

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

At 1920x1080, the measured average frame rates are as follows: Low settings produce an average of 203 FPS, Medium produces 144 FPS with a minimum of 122 FPS and a maximum of 165 FPS, High produces 128 FPS, and Ultra produces 111 FPS. The spread between Low and Ultra is 92 FPS, indicating a wide performance range that allows users to prioritize frame rate or fidelity.

At 2560x1440, the averages are: Low at 153 FPS, Medium at 109 FPS with a minimum of 93 FPS and a maximum of 125 FPS, High at 97 FPS, and Ultra at 84 FPS. The delta between Low and Ultra here is 69 FPS, showing a narrower range than at 1080p, which is expected as the GPU becomes more stressed.

At 3840x2160, the averages are: Low at 103 FPS, Medium at 73 FPS with a minimum of 62 FPS and a maximum of 85 FPS, High at 65 FPS, and Ultra at 57 FPS. The spread at 4K is just 46 FPS, demonstrating that the GPU is the limiting factor. The minimum FPS data for Medium settings at each resolution (122 at 1080p, 93 at 1440p, 62 at 4K) shows a consistent trend: the frame pacing remains stable, with the minimum never falling more than 15% below the average.

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

The frame rate scaling from 1920x1080 to 3840x2160 reveals a classic GPU-bound profile. For the High preset, the average FPS drops from 128 at 1080p to 97 at 1440p (a 24.2% decrease) and then to 65 at 4K (a 33.0% decrease from 1440p). The total drop from 1080p to 4K is 63 FPS, or a 49.2% reduction. For the Medium preset, the averages are 144 at 1080p, 109 at 1440p (a 24.3% decrease), and 73 at 4K (a 33.0% decrease from 1440p), for a total reduction of 71 FPS (49.3%). The Low preset shows a similar pattern: 203 at 1080p, 153 at 1440p (a 24.6% decrease), and 103 at 4K (a 32.7% decrease from 1440p), totaling 100 FPS lost (49.3%).

The consistency of these percentage drops across all presets — approximately 24% from 1080p to 1440p and 33% from 1440p to 4K — is a strong indicator that the GPU is the primary bottleneck at resolutions above 1080p. If the CPU were the limiting factor, the percentage drop from 1080p to 1440p would be much smaller, because the CPU would still be able to feed the GPU at 1440p, and the FPS would remain nearly flat until the resolution became too high for the GPU. Instead, the linear scaling of frame rates with pixel count (4K has 2.25 times the pixels of 1440p, and 1440p has 1.78 times the pixels of 1080p) points to the GPU’s fill rate and shading units being the constraining resource. The fact that the 1080p-to-1440p drop is slightly smaller (24%) than the 1440p-to-4K drop (33%) suggests that at 1080p, the CPU is starting to limit performance slightly, preventing the GPU from reaching its full potential. This is corroborated by the 203 FPS Low preset at 1080p, which is high but not as high as the GPU’s raw compute would suggest if the CPU could keep up perfectly. The data overall indicates that for the best experience, users should target 1440p or 4K with Medium settings to avoid the CPU bottleneck that begins to appear at 1080p.