Microsoft Flight Simulator

The AMD Ryzen 9 9950X3D and Intel Arc B390 pairing presents one of the most lopsided combinations in the Microsoft Flight Simulator benchmark database. The CPU is a 16-core, 32-thread Zen 5 part with a base clock of 4.30 GHz and a boost clock of 5.70 GHz, built on a 4 nm TSMC process with 128 MB of L3 cache. The GPU, in contrast, is an integrated graphics solution based on the Xe3-LPG architecture with 1536 shading units, 48 texture mapping units, and 24 raster operation pipelines. Benchmark results show this combo ranks 2665 out of 2996 tested combinations, placing it in the bottom tier despite the processor’s top-tier compute capabilities. The measured frame rates across resolutions and settings tell a clear story of GPU-bound performance, where the Arc B390’s capabilities—not the 9950X3D’s—dictate the experience.

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

The Ryzen 9 9950X3D brings substantial processing power to the table. Its 16 cores and 32 threads operate with a base clock of 4.30 GHz and a boost clock of 5.70 GHz, backed by a 128 MB L3 cache. The CPU’s benchmark scores reinforce this strength: Cinebench R23 multicore reaches 42018 points, while the single-core score is 2259. Geekbench results show 26736 multicore and 3064 single-core. In Passmark, the multithread score is 70255, with single-thread at 4737. The 3DMark tests show 17184 for max threads, 16374 for 16 threads, and 1292 for single-thread. These numbers place the CPU at the 95th percentile among all processors, with an average benchmark score of 75779.

In Microsoft Flight Simulator, the CPU’s role is substantial during terrain loading, physics calculations, and complex airport scenes. The 9950X3D’s high single-thread performance—evidenced by the 2259 Cinebench R23 single-core score—helps maintain frame pacing in scenarios where the simulation logic is the bottleneck. However, the measured FPS data shows that this CPU is rarely the limiting factor in this combo. At 1080p Low, the average FPS is 56, which is the highest measured across all settings and resolutions. The fact that lowering resolution from 4K to 1080p at Low settings improves average FPS from 33 to 56—a 70% increase—indicates that the GPU, not the CPU, is the constraint. The 9950X3D has ample headroom; its nearest rivals include the AMD Ryzen 7 PRO 9755 with an average score of 75738 (0.1% delta), the AMD Ryzen 7 PRO 9755X3D at 75716 (0.1% delta), and the AMD EPYC 8224P at 75582 (0.3% delta). The Intel Core Ultra 9 285 scores 75488, a 0.4% delta. This CPU is effectively at the top of its class, yet the combo’s rank of 2665 out of 2996 shows it cannot compensate for the GPU’s limitations.

The CPU’s 128 MB L3 cache is particularly relevant for flight simulators, which often stream large amounts of terrain and object data. The 16,630 million transistors on a 2x 70.6 mm² die size provide the computational headroom needed for complex flight models. However, benchmark results indicate that even this powerful CPU cannot push frame rates past the GPU’s ceiling. The data suggests that in this game, the 9950X3D’s role is to prevent CPU-side bottlenecks, which it does effectively—the FPS differences between settings at the same resolution are consistent with GPU workload scaling, not CPU limitations.

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

The Intel Arc B390 is an integrated graphics processor with a base clock of 300 MHz and a boost clock of 2500 MHz. Its memory configuration is system-shared, meaning it relies on the host system’s DDR5 RAM rather than dedicated VRAM. The GPU’s compute capabilities are modest: FP32 performance is 7.680 TFLOPS, with FP16 at 15.36 TFLOPS (2:1 ratio). The pixel rate is 60.00 GPixel/s, and the texture rate is 120.0 GTexel/s. The GPU supports DirectX 12 Ultimate (12_2), OpenGL 4.6, and Vulkan 1.4, but its performance percentile is only 9th among all GPUs, with an average benchmark score of 1482 in 3DMark Steel Nomad DX12.

The GPU’s nearest rivals in the database highlight its position: the NVIDIA GeForce GT 520MX scores 1463 (1.3% delta), the NVIDIA GeForce 800M scores 1460 (1.5% delta), and the NVIDIA GeForce GT 625 OEM scores 1446 (2.5% delta). These are entry-level parts from previous generations, indicating the Arc B390 is not designed for demanding 3D workloads. In Microsoft Flight Simulator, this GPU becomes the primary bottleneck. The system-shared memory bandwidth is system-dependent, which means performance can vary based on the host platform, but the measured results show consistent limitations.

At 4K Ultra, the average FPS is 19, and at 4K High it is also 19. Dropping to 4K Low improves to 33 FPS, but that is still far below playable standards for most users. The 2500 MHz boost clock does not compensate for the limited shading units and memory bandwidth. The GPU’s 24 ROPs and 48 TMUs are insufficient for the heavy geometry and texture loads in flight simulation. The data shows that regardless of the CPU’s power, the Arc B390 caps performance at levels that are roughly half of what a dedicated mid-range GPU would achieve. The 3DMark Steel Nomad score of 1482, compared to the CPU’s 3DMark max threads score of 17184, illustrates the massive gap between the two components’ capabilities.

How This Combo Ranks — use comboRankInGame vs other tested combos

The combo rank of 2665 out of 2996 tested combinations places this pairing in the bottom 11% of all tested systems for Microsoft Flight Simulator. This rank is particularly striking given the CPU’s 95th percentile standing among processors. The discrepancy between CPU and GPU performance classes is the defining characteristic of this combo. The 9950X3D is a flagship desktop processor with a 170 W TDP, while the Arc B390 is an integrated GPU with an 80 W TDP, designed for power efficiency rather than raw performance.

Benchmark results indicate that the combo’s rank is determined almost entirely by the GPU. The CPU’s nearest rivals—all within 0.4% of its average score—would not meaningfully change the FPS results if paired with the Arc B390. Conversely, replacing the Arc B390 with even a modest dedicated GPU would dramatically improve the rank. The data shows that this combo is not balanced; the CPU is overqualified for the GPU’s output. In practical terms, a user would see better results by pairing the 9950X3D with a stronger GPU, but the measured data only reflects this specific combination.

The rank of 2665 means that 331 tested combos perform worse. These are likely systems with both weak CPUs and weak GPUs, whereas this combo has an extremely strong CPU paired with a weak GPU. The ranking is a composite of both components’ contributions, and the GPU’s 9th percentile performance drags the combo down despite the CPU’s excellence. The delta between the CPU’s percentile (95) and the combo’s rank (bottom 11%) quantifies how much the GPU limits overall performance.

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

The measured FPS data provides a full picture of this combo’s performance across three resolutions and four settings levels. At 1920x1080, the average FPS ranges from 56 at Low settings to 31 at Ultra. The Low setting achieves 56 FPS, which is the best result in the entire dataset. Medium settings produce 43 FPS, with minimum 37 and maximum 50. High settings yield 33 FPS, and Ultra drops to 31 FPS. The progression from Low to Ultra at 1080p shows a 45% reduction in average FPS, from 56 to 31.

At 2560x1440, the results scale downward. Low settings average 45 FPS, Medium averages 35 FPS (minimum 30, maximum 40), High averages 26 FPS, and Ultra averages 25 FPS. The drop from 1080p Low (56 FPS) to 1440p Low (45 FPS) represents a 20% decrease, while the drop from 1080p Ultra (31 FPS) to 1440p Ultra (25 FPS) is a 19% decrease. This consistency suggests the GPU is equally stressed across settings when resolution increases.

At 3840x2160, the results are marginal. Low settings average 33 FPS, Medium averages 26 FPS (minimum 22, maximum 30), High averages 19 FPS, and Ultra averages 19 FPS. The 4K Low result of 33 FPS is identical to the 1080p High result, showing that resolution scaling has a more significant impact than settings changes. The minimum FPS of 22 at 4K Medium indicates that even the lower-bound performance is below playable thresholds. The 4K results show that this combo cannot maintain smooth gameplay at any setting, with the highest average being 33 FPS at Low.

The data reveals that the gap between Low and Ultra settings narrows as resolution increases. At 1080p, the difference between Low (56) and Ultra (31) is 25 FPS. At 1440p, the difference is 20 FPS (45 to 25). At 4K, the difference is 14 FPS (33 to 19). This narrowing indicates that at higher resolutions, the GPU is fully saturated, and settings changes have diminishing returns.

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

Resolution scaling in this combo shows a clear GPU-bound pattern. From 1080p Low (56 FPS) to 1440p Low (45 FPS), the drop is 11 FPS, or about 20%. From 1440p Low to 4K Low (33 FPS), the drop is another 12 FPS, or about 27%. The cumulative drop from 1080p Low to 4K Low is 23 FPS, a 41% reduction. This scaling is consistent with a GPU that has limited fill rate and memory bandwidth—as pixel count increases, the GPU must work proportionally harder, and the system-shared memory becomes a bottleneck.

At High settings, the scaling is steeper in percentage terms. 1080p High averages 33 FPS, 1440p High averages 26 FPS (a 21% drop), and 4K High averages 19 FPS (a 27% drop from 1440p). The total drop from 1080p High to 4K High is 14 FPS, or 42%. At Ultra settings, 1080p Ultra (31 FPS) falls to 1440p Ultra (25 FPS, a 19% drop) and then to 4K Ultra (19 FPS, a 24% drop). The total drop is 12 FPS, or 39%.

The percentage drops are similar across settings, which indicates that the GPU is the limiting component at all settings. If the CPU were the bottleneck, we would expect to see smaller FPS drops at lower resolutions (where the CPU has more influence) and larger drops at higher resolutions (where the GPU takes over). Instead, the consistent ~20% drop per resolution step suggests the GPU is always saturated. The CPU’s power is untapped; the 9950X3D could easily handle higher frame rates if the GPU could deliver them.

The data also shows that 4K performance is essentially halved compared to 1080p at every setting. This halving is typical for a GPU with limited memory bandwidth, as the system-shared memory cannot keep up with the increased data throughput required at higher resolutions. The 4K results, with averages between 19 and 33 FPS, are below the threshold for smooth flight simulation, which typically requires at least 30 FPS for acceptable control response.

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

Based on the measured FPS data, the optimal settings depend on the user’s resolution and tolerance for frame rate drops. At 1920x1080, the Low preset provides the best experience with an average of 56 FPS. This is the only setting that achieves a consistently playable frame rate, as Medium (43 FPS) is also playable but with a 23% reduction. High (33 FPS) and Ultra (31 FPS) are marginal, with averages below the 35 FPS threshold that many flight sim enthusiasts prefer for smooth panning.

At 2560x1440, the Low preset averages 45 FPS, which is the only setting that remains above 40 FPS. Medium averages 35 FPS, with a minimum of 30 FPS, which is borderline playable but may experience stutters in dense scenery. High (26 FPS) and Ultra (25 FPS) are not recommended, as they fall below 30 FPS. The data suggests that 1440p users should stick to Low or Medium settings, accepting the visual trade-offs for smoother performance.

At 3840x2160, no setting provides a fully playable experience. The Low preset averages 33 FPS, which is technically above 30 but may still feel sluggish in complex scenes. Medium averages 26 FPS, with a minimum of 22 FPS, which is likely to cause noticeable judder. High and Ultra both average 19 FPS, which is not viable for interactive flight simulation. For 4K users, the recommendation is to lower the resolution to 1440p or 1080p, as the GPU cannot handle 4K at any setting.

The best overall experience in the measured data is 1080p Low, with 56 FPS average. This provides the smoothest frame pacing and the most headroom for sudden drops in dense areas. For users who prioritize visual quality over frame rate, 1080p Medium at 43 FPS offers a reasonable compromise, with minimum 37 FPS indicating relatively stable performance. The Ultra preset at 1080p (31 FPS) is not worth the visual gain over High (33 FPS), given the minimal FPS difference.

The data shows that this combo is best suited for 1080p gaming with Low or Medium settings. At higher resolutions, the GPU’s limitations become prohibitive. Users with this combo should consider reducing their display resolution or upgrading the GPU, as the CPU is not the limiting factor. The measured results indicate that even at 4K Low, the FPS (33) is lower than 1080p High (33), meaning that resolution has a greater impact on performance than settings. This insight suggests that users should prioritize resolution reduction before adjusting quality presets.