Control

The AMD Ryzen 9 5950X paired with the Intel Arc B390 in Control produces a deeply lopsided result. The data shows a 16-core, 32-thread Zen 3 processor from the 5000 series, with a base clock of 3.40 GHz and a boost clock of 4.90 GHz, driving an integrated-class GPU that shares system memory. This is not a balanced pairing; the benchmark results reflect a system where the graphics solution is the overwhelming constraint at every measured setting and resolution. The measured FPS figures are consistently low, and the scaling patterns between resolutions and settings point directly at the GPU as the limiting factor. For a builder considering this combination, the practical takeaway is straightforward: the CPU’s immense compute headroom is irrelevant to Control’s frame rate, as the Arc B390’s output caps the experience well below playable thresholds in most configurations.

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

The Ryzen 9 5950X is a desktop processor built on TSMC’s 7 nm process, with a Zen 3 architecture codenamed Vermeer. It features 16 cores and 32 threads, backed by 64 MB of L3 cache and a dual-channel DDR4 memory bus with 51.2 GB/s of bandwidth. Its benchmark scores are substantial: Cinebench R23 multicore reaches 26017, while the single-core test scores 1614. Geekbench multicore sits at 15233, and single-core at 2083. The processor also delivers a Passmark multithread score of 45424 and a single-thread score of 3470. These numbers place the 5950X in the 91st percentile of all CPUs, with an average benchmark score of 51947. Its nearest rivals include the Intel Core Ultra 5 235HX (average score 52073, delta -0.2%), AMD EPYC 8124P (52121, delta -0.3%), Intel Core i9-13900F (51730, delta 0.4%), and Intel Core i7-14700 (52301, delta -0.7%). The deltas are negligible, meaning the 5950X trades blows with modern high-end parts within a fraction of a percent.

In Control, however, the CPU’s role is minimal. The measured FPS at 1920x1080 with Low settings is 65 avg, which is the highest figure in the entire dataset. At 1080p Ultra, the average drops to 36. At 3840x2160 Ultra, it falls to 18. These numbers do not reflect a CPU bottleneck; a processor capable of 26017 in Cinebench R23 multicore would not be the reason a game runs at 21 FPS at 4K High. The 5950X has more than enough single-thread and multi-thread throughput to handle Control’s logic, physics, and draw calls at any of the tested settings. The data supports this: moving from 1080p Low (65 FPS) to 1080p Ultra (36 FPS) reduces performance by 44.6%, which is a settings-driven drop, not a CPU limitation. The CPU’s 3.40 GHz base and 4.90 GHz boost clocks are ample for the game’s requirements. The fact that the 5950X’s 3DMark scores (e.g., 10810 for 16 threads, 11597 for max threads) are high reinforces that the processor is not the weak link. In this combo, the CPU is effectively idle in terms of frame rate contribution; its job is to feed frames to a GPU that cannot keep up.

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

The measured FPS rows provide a clear hierarchy. At 1920x1080, the Low preset yields an average of 65 FPS, which is the only setting that approaches smooth gameplay. Medium drops to 46 avg, High to 41, and Ultra to 36. At 2560x1440, Low averages 50 FPS, Medium 35, High 31, and Ultra 27. At 3840x2160, Low averages 33, Medium 24, High 21, and Ultra 18. The pattern is consistent: lowering settings increases FPS, but the gains vary by resolution. At 1080p, moving from Ultra to Low improves average FPS by 29 points (36 to 65), a 80.6% increase. At 1440p, the same move improves by 23 points (27 to 50), a 85.2% increase. At 4K, the improvement is 15 points (18 to 33), an 83.3% increase.

For the best experience, the data suggests 1080p Low is the only viable preset for consistent playability, as it is the sole configuration above 60 FPS. Medium at 1080p (46 avg) and Low at 1440p (50 avg) are borderline, but the min FPS figures are only available for Medium presets: at 1080p Medium, min is 40; at 1440p Medium, min is 30; at 4K Medium, min is 20. Low presets have no min FPS recorded, so stability is unknown, but the averages are higher. If a user prioritizes frame rate, 1080p Low is the recommendation. If visual quality is more important, the data shows that 1080p Ultra (36 avg) is playable but not smooth, and 1440p Medium (35 avg) offers a similar average with better resolution. The Ultra preset at 1080p gives a 36 avg, which is 55.4% of the Low preset’s 65 avg. No setting at 4K achieves more than 33 FPS, so that resolution should be avoided for this combo. The measured rows indicate that the GPU’s performance ceiling is low, so the best experience is found at the lowest resolution and settings.

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

The FPS drop from 1080p to 4K is severe and consistent across all settings. At Low, the average goes from 65 FPS at 1080p to 50 at 1440p (a 23.1% drop) and then to 33 at 4K (a 34% drop from 1440p). At Medium, the sequence is 46, 35, and 24, with drops of 23.9% and 31.4%. At High, it is 41, 31, and 21, with drops of 24.4% and 32.3%. At Ultra, it is 36, 27, and 18, with drops of 25% and 33.3%. The percentage drops are remarkably consistent, which indicates a uniform scaling penalty tied to pixel count. The resolution change from 1080p to 4K represents a 4x increase in pixels, yet the FPS does not drop by 4x; it drops by roughly half to two-thirds depending on the setting. For example, at Low, 65 FPS at 1080p becomes 33 FPS at 4K, a 49.2% reduction. At Ultra, 36 becomes 18, a 50% reduction. This non-linear scaling suggests that the GPU is not purely pixel-bound; other factors like memory bandwidth (system shared) and the GPU’s architecture play a role.

The consistency of the scaling across settings is telling. If the CPU were the limiter, we would expect FPS to stay flat or drop less at higher resolutions, because the CPU load would remain similar while the GPU load increases. Instead, the FPS drops proportionally with resolution, which is the classic signature of a GPU bottleneck. The Intel Arc B390 has a boost clock of 2500 MHz, a base clock of 300 MHz, and 1536 shading units. Its 3DMark Steel Nomad DX12 score is 1482, which places it in the 9th percentile of all GPUs. Its nearest rivals are the NVIDIA GeForce GT 520MX (avg score 1463, delta 1.3%), NVIDIA GeForce 800M (1460, delta 1.5%), NVIDIA GeForce GT 625 OEM (1446, delta 2.5%), and NVIDIA GeForce GT 710 (1443, delta 2.7%). These are all ancient or entry-level parts, confirming the Arc B390’s low-end standing. The resolution scaling data aligns with this: the GPU cannot handle the pixel throughput demands of 4K, so FPS collapses. The limiting component is unequivocally the GPU.

How This Combo Ranks — use comboRankInGame vs other tested combos

The combo’s rank in Control is 2551 out of 2888 tested combinations. This places it in the bottom 11.7% of all tested systems, meaning 2550 combos performed better, and only 337 performed worse. This low rank is expected given the GPU’s 9th percentile standing. The Ryzen 9 5950X is a high-end CPU, but the combo’s performance is dragged down by the Arc B390. In the context of Control, the game’s requirements are modest for CPUs but demanding on GPUs, so a weak GPU with a strong CPU will rank poorly. The delta between this combo and the top performers is not quantified in the data, but the rank alone tells the story: this is a bottom-tier combination for this game.

Comparing to the CPU’s rivals does not help the combo’s case. The 5950X’s nearest rivals (Intel Core Ultra 5 235HX, AMD EPYC 8124P, Intel Core i9-13900F, Intel Core i7-14700) all have similar average CPU scores, but none of them are paired with an Arc B390 in this dataset. The GPU’s rivals (GT 520MX, 800M, GT 625 OEM, GT 710) are all similarly weak, with deltas under 3%. This means that swapping the Arc B390 for any of those GPUs would not meaningfully change the combo’s rank. The bottleneck is so severe that the CPU’s power is wasted. For a builder, this rank is a warning: the 5950X deserves a far more capable GPU to avoid being in the bottom percentile. The combo’s position at 2551 out of 2888 is a direct consequence of the GPU’s performance ceiling, and no CPU-side adjustment can fix it.

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

The Intel Arc B390 is an integrated GPU (IGP) based on the Xe3-LPG architecture, built on Intel’s 3 nm process. Its memory is System Shared, meaning it uses the system’s DDR4 RAM with bandwidth described as System Dependent. The GPU has no dedicated VRAM, which is a critical weakness for a game like Control that uses high-resolution textures and effects. The base clock is 300 MHz, and the boost clock is 2500 MHz. It has 1536 shading units, 48 TMUs, and 24 ROPs, along with 12 ray tracing cores. The pixel rate is 60.00 GPixel/s, and the texture rate is 120.0 GTexel/s. FP32 performance is 7.680 TFLOPS, and FP16 is 15.36 TFLOPS (2:1). The TDP is 80 W, and it uses no power connectors, confirming its integrated nature.

The GPU’s 3DMark Steel Nomad DX12 score is 1482, with a percentile of 9. This is an extremely low score. Its nearest rivals are all NVIDIA parts from over a decade ago, with the GT 710 being a common passive-cooled card for basic displays. The delta between the Arc B390 and the GT 710 is 2.7%, meaning the Arc B390 is only slightly faster. In Control, this GPU produces an average of 65 FPS at 1080p Low, but that is the best case. At 4K Ultra, it produces 18 FPS. The system-shared memory is a major factor: without dedicated VRAM, the GPU competes with the CPU for memory bandwidth, and the 51.2 GB/s of the 5950X’s dual-channel DDR4 is far below what a discrete GPU would use. The boost clock of 2500 MHz is decent, but the low ROP count (24) and limited shading units (1536) cap the fill rate. The measured FPS figures directly reflect these specifications: the GPU is only capable of 33 FPS at 4K Low, which is unplayable. The GPU role is that of a bottleneck; every frame rate in the dataset is limited by its compute and memory constraints, not by the CPU or game code.

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

The measured FPS data covers three resolutions and four settings each. At 1920x1080, the Low preset averages 65 FPS, which is the highest score in the entire dataset. Medium averages 46 FPS with a min of 40 and max of 53. High averages 41 FPS. Ultra averages 36 FPS. The progression from Low to Ultra shows a steady decline: 65, 46, 41, 36. The gap between Low and Medium is 19 FPS (a 29.2% drop), while the gap between Medium and High is 5 FPS (a 10.9% drop), and between High and Ultra is 5 FPS (a 12.2% drop). This suggests diminishing returns: the first step from Low to Medium is the most impactful, while subsequent steps cost less performance.

At 2560x1440, the Low preset averages 50 FPS. Medium averages 35 FPS with a min of 30 and max of 40. High averages 31 FPS. Ultra averages 27 FPS. The progression is 50, 35, 31, 27. The drop from Low to Medium is 15 FPS (30%), from Medium to High is 4 FPS (11.4%), and from High to Ultra is 4 FPS (12.9%). Compared to 1080p, the Low preset drops from 65 to 50, a 23.1% reduction. The Medium preset drops from 46 to 35, a 23.9% reduction. High drops from 41 to 31, a 24.4% reduction. Ultra drops from 36 to 27, a 25% reduction.

At 3840x2160, the Low preset averages 33 FPS. Medium averages 24 FPS with a min of 20 and max of 27. High averages 21 FPS. Ultra averages 18 FPS. The progression is 33, 24, 21, 18. The drop from Low to Medium is 9 FPS (27.3%), from Medium to High is 3 FPS (12.5%), and from High to Ultra is 3 FPS (14.3%). The resolution scaling from 1440p to 4K shows Low dropping from 50 to 33 (34%), Medium from 35 to 24 (31.4%), High from 31 to 21 (32.3%), and Ultra from 27 to 18 (33.3%). The min FPS values are only recorded for Medium presets: at 1080p it is 40, at 1440p it is 30, and at 4K it is 20. These mins are 87%, 85.7%, and 83.3% of their respective averages, indicating some frame pacing variance but not severe stuttering. Across all resolutions, the Low preset is the only one that exceeds 50 FPS at any resolution except 4K (33 FPS). The data shows that the combo’s playable envelope is limited to 1080p Low and 1440p Low, with everything else falling below 50 FPS. The Ultra preset never exceeds 36 FPS at any resolution. This breakdown confirms that the Intel Arc B390 is the definitive limiter, and the Ryzen 9 5950X’s capabilities are never utilized.