Control

The Intel Core i5-10400F paired with the NVIDIA GeForce RTX 5090 presents a striking contrast in a title like Control, where the GPU's raw power is immense but the CPU's age shows in frame pacing. The measured data across resolutions reveals that the most balanced experience is achieved at the High preset. At 1920x1080, High delivers 145.1 FPS, which is only 24.5 FPS behind the Low preset’s maximum of 169.6 FPS, but it offers substantially better visual fidelity. Moving to 2560x1440, High produces 132.7 FPS, a figure that remains highly fluid and only 18.7 FPS behind Low. The gap narrows further at 3840x2160, where High yields 90.2 FPS against Low’s 137.2 FPS. While Ultra provides the highest image quality, its performance cost is steep: it drops to 135.8 FPS at 1080p, 115.6 FPS at 1440p, and 81 FPS at 4K. The data indicates that High is the sweet spot, as it maintains a comfortable margin above 90 FPS at all resolutions, whereas Medium offers only marginal gains over High (148 FPS vs. 145.1 FPS at 1080p) with less visual polish.

CPU Role

The Intel Core i5-10400F is a 6-core, 12-thread processor from the Comet Lake architecture, with a base clock of 2.90 GHz and a boost clock of 4.30 GHz. Its L3 cache is 12 MB shared, and it supports DDR4 memory with a dual-channel bus. In synthetic benchmarks, this CPU scores 4,748 in 3DMark with 16 threads, 2,560 with 4 threads, and 688 in single-thread tests. Its Cinebench R23 results are 10,297 multi-core and 1,453 single-core. The processor’s percentile versus all CPUs is 73, indicating it sits above the median but is not a high-end part.

In Control, this CPU’s role is primarily to feed the RTX 5090. At 1080p, the data shows a clear ceiling: the average FPS only scales from 169.6 (Low) to 135.8 (Ultra), a 33.8 FPS spread. This narrow range suggests that the game’s engine is heavily reliant on the CPU’s single-thread performance at lower resolutions, as the GPU has ample headroom. The 3DMark single-thread score of 688 and the Cinebench R23 single-core score of 1,453 are modest, and they align with the observed behavior where the CPU limits absolute frame rates. At 1440p, the spread narrows further (151.4 to 115.6), indicating the GPU begins to take on more load. The CPU’s 12 threads do help maintain consistent minimums, as the game’s physics and AI operations are parallelizable, but the 4.30 GHz boost clock is the key limiter. The benchmark results show that this CPU is adequate for 60+ FPS gameplay, but it cannot push beyond the 170 FPS threshold even at Low settings, a direct consequence of its 6-core/12-thread configuration and 14 nm process node.

GPU Role

The NVIDIA GeForce RTX 5090 is a Blackwell 2.0 architecture GPU with 21,760 shading units, 680 TMUs, and 176 ROPs. It features 170 RT cores and 680 tensor cores. Its memory subsystem is substantial: 32 GB of GDDR7 on a 512-bit bus, yielding a bandwidth of 1.79 TB/s. The GPU’s clocks are 2017 MHz base and 2407 MHz boost, with memory running at 1750 MHz. In synthetic tests, it scores 14,411 in 3DMark Steel Nomad DX12, 380,114 in Geekbench OpenCL, and 39,650 in Passmark G3D. Its percentile versus all GPUs is 94, placing it in the top tier.

In Control, the GPU’s massive VRAM pool is largely irrelevant, as the game’s textures are not particularly demanding. However, the core compute power drives the resolution scaling. At 4K, the RTX 5090 delivers 90.2 FPS on High, 101.2 FPS on Medium, and 137.2 FPS on Low. The drop from 1080p High (145.1 FPS) to 4K High (90.2 FPS) is a 54.9 FPS reduction, which is a 37.8% performance loss. This is a significant scaling factor, indicating that the GPU is the primary bottleneck at 4K. The 104.8 TFLOPS FP32 compute and 423.6 GPixel/s pixel rate are clearly being utilized, as the GPU’s fill rate and shading capacity directly affect the game’s rendering of particle effects and lighting. The RTX 5090’s 2407 MHz boost clock ensures high throughput, but the data suggests that even this flagship GPU cannot maintain 100+ FPS at 4K Ultra, capping at 81 FPS. The GPU’s 1.79 TB/s bandwidth is sufficient to avoid memory bottlenecks, as the 32 GB pool is never taxed. The measured results show that the GPU is the dominant component at 4K, while at 1080p, its power is held back by the CPU.

How This Combo Ranks

The combination of the Core i5-10400F and RTX 5090 ranks 145th out of 1,636 tested combos in Control, placing it in the top 8.9% of all configurations. This is a strong result, but it is notably lower than what the GPU alone would suggest, given its 94th percentile ranking. The CPU’s 73rd percentile ranking drags the overall score down, as the system is clearly imbalanced. The data shows that the combo’s performance is closer to what a mid-range CPU with a top-tier GPU would achieve, rather than a full flagship pairing. In the nearest rivals for the CPU, the AMD EPYC 7552 is only 0.4% ahead in average benchmark score, and the Intel Core i5-8500 is 1.1% behind, indicating that the 10400F is a competent but not exceptional processor. For the GPU, the nearest rival is the RTX 5090 D, which is 0.1% ahead in average score, showing that the 5090 is at the very top of its class.

Given this rank, the combo is positioned well for high-refresh-rate 1440p gaming, but it underperforms at 1080p where the CPU bottleneck is most apparent. The 145th rank out of 1,636 suggests that many other combos, likely with newer CPUs, achieve higher average FPS. The delta between the CPU’s rivals (0.4% to -1.2%) is small, meaning the 10400F is not a significant outlier, but the GPU’s rivals (0.1% to -1.6%) are also close, indicating that the GPU is performing as expected. The overall rank is thus primarily a function of the CPU’s limitations, as the GPU’s score is near the top of the distribution. In practical terms, this combo will deliver excellent frame rates, but it will leave performance on the table in CPU-bound scenes.

Resolution Scaling

The measured FPS data reveals a clear pattern of resolution scaling that shifts the bottleneck from the CPU to the GPU. At 1920x1080, the average FPS across all settings ranges from 169.6 (Low) to 135.8 (Ultra), a spread of 33.8 FPS. The difference between Low and High is 24.5 FPS, and between High and Ultra is 9.3 FPS. This tight clustering indicates that the CPU is the limiting factor, as the GPU is capable of much higher frame rates but is constrained by the processor’s ability to issue draw calls and update game logic. The 3DMark 16-thread score of 4,748 and the Cinebench R23 multi-core score of 10,297 are the relevant metrics here, as they show the CPU’s parallel throughput is moderate, but its single-thread performance (688 in 3DMark single-thread) is the bottleneck.

Moving to 2560x1440, the spread widens to 35.8 FPS (151.4 Low to 115.6 Ultra), but the absolute values drop. The High preset yields 132.7 FPS, which is a 12.4 FPS reduction from 1080p High. This 8.5% drop is modest, suggesting that the GPU is starting to take on more load, but the CPU is still holding the system back. At 3840x2160, the scaling becomes pronounced. The Low preset drops to 137.2 FPS, which is only 32.4 FPS below the 1080p Low figure, indicating a significant GPU workload increase. The High preset falls to 90.2 FPS, a 54.9 FPS drop from 1080p High, which is a 37.8% reduction. The Ultra preset drops to 81 FPS, a 54.8 FPS reduction from 1080p Ultra. This 4K data shows that the GPU is now the dominant limiter, as the RTX 5090’s 104.8 TFLOPS compute is being saturated by the higher pixel count.

The overall trend shows that the CPU bottleneck is most severe at 1080p, where the FPS difference between Low and Ultra is only 33.8 FPS, but it becomes less relevant at 4K, where the difference is 56.2 FPS (137.2 Low to 81 Ultra). This indicates that for users targeting 1080p high-refresh monitors, the 10400F will cap performance, while at 4K, the RTX 5090’s raw power is the deciding factor. The data suggests that this combo is best suited for 1440p gaming, where the balance is more equitable, but at 4K, the system delivers playable frame rates across all settings, with High being the recommended preset for a smooth 90 FPS experience.