Microsoft Flight Simulator

# How This Combo Ranks

The AMD Ryzen 9 9900X paired with the Intel Arc B370 lands at rank 2770 out of 2996 tested combos in Microsoft Flight Simulator. That places this configuration in the bottom 8% of all tested combinations, a result that reflects the severe bottleneck introduced by the integrated-class GPU. The combo sits far below the median performer, and the data indicates that the CPU's considerable processing power is largely wasted in this pairing. The rank is particularly striking given that the Ryzen 9 9900X itself sits in the 92nd percentile among all CPUs, while the Arc B370 languishes in the 5th percentile among all GPUs. This is a textbook case of a high-end processor shackled to a graphics solution that cannot keep pace with even modest rendering demands.

The delta between the two components' relative standings is the story here. The CPU's average benchmark score of 57498 places it within 1.4% of much larger enterprise silicon like the Intel Xeon Platinum 8260M, and it actually edges out the AMD EPYC 7313 by 0.2%. Meanwhile, the GPU's average score of 1184 puts it on par with legacy mobile parts like the ATI Mobility Radeon HD 5570, which trails by just 0.2%. The combination of a top-decile processor with a bottom-decile GPU produces a composite rank that neither component would individually suggest. The measured FPS data across all resolutions and settings confirms that the Arc B370 is the limiting factor, capping performance at levels that even the weakest desktop CPUs could achieve with a proper discrete GPU.

GPU Role

The Intel Arc B370 is an integrated graphics solution built on the Xe3-LPG architecture, derived from the Panther Lake chip. Its specifications reveal why it struggles in a demanding flight simulator: the GPU operates with system-shared memory, meaning there is no dedicated VRAM pool. The memory bus width and bandwidth are both listed as "System Shared" and "System Dependent," respectively, which means performance scales with the host system's DDR5 memory configuration rather than any fixed graphics memory design. The base clock sits at 300 MHz with a boost of 2400 MHz, but these clocks cannot compensate for the fundamental bandwidth limitations of shared memory.

The GPU's compute resources are minimal by modern standards. It offers 1280 shading units, 40 texture mapping units, and 20 raster operation units, along with 10 ray tracing cores. The pixel rate is 48.00 GPixel/s and the texture rate is 96.00 GTexel/s, figures that are dwarfed by even entry-level discrete GPUs of the past decade. Floating point performance is rated at 6.144 TFLOPS for FP32 and 12.29 TFLOPS for FP16, but these numbers do not translate into playable frame rates in a simulation that stresses both geometry and texture bandwidth. The 3DMark Steel Nomad DX12 score of 1184 places this GPU in the 5th percentile of all GPUs, and its nearest rivals include the ATI Radeon HD 5770 (scoring 1190, just 0.5% higher) and the AMD FirePro M2000 (scoring 1168, which the Arc B370 beats by 1.4%).

In Microsoft Flight Simulator specifically, the GPU's role is to render detailed terrain, aircraft models, and dynamic weather. The data shows that even at 1080p Low settings, the average FPS is just 50, and at 4K Ultra it drops to 17. These numbers indicate that the GPU is overwhelmed by the game's rendering workload. The shared memory architecture means that the 12-core, 24-thread Ryzen 9 9900X must also feed system memory to the GPU, creating contention that further reduces performance. The GPU's power envelope of 25 W (as an IGP) and lack of dedicated power connectors underscore its positioning as a basic display solution, not a gaming part.

Resolution Scaling

The FPS scaling across resolutions reveals a GPU that is saturated even at the lowest tested resolution. Moving from 1080p to 1440p at Low settings, the average FPS drops from 50 to 41, a decrease of 18%. From 1440p Low to 4K Low, the average falls from 41 to 30, another 27% reduction. This scaling pattern is consistent with a GPU that is compute-bound rather than bandwidth-bound, as the percentage drops are relatively modest compared to what a discrete GPU might show when transitioning between these resolutions.

At High settings, the scaling is steeper. The 1080p High average is 30 FPS, which drops to 24 FPS at 1440p High (a 20% reduction) and then to 18 FPS at 4K High (a further 25% reduction). The Ultra preset shows a similar trend: 28 FPS at 1080p, 23 FPS at 1440p, and 17 FPS at 4K. The relatively small absolute differences between resolutions—often just 5-7 FPS per step—indicate that the limiting component is not the memory bandwidth (which would show larger penalties at 4K) but rather the raw shading and texture throughput of the GPU. The Arc B370's 96.00 GTexel/s texture rate and 48.00 GPixel/s pixel rate are simply insufficient for the game's high-resolution textures, regardless of the resolution selected.

The data suggests that the CPU is not the bottleneck at any resolution. If the Ryzen 9 9900X were limiting performance, we would expect to see larger FPS increases when lowering resolution, as the GPU would have more headroom to render frames. Instead, the FPS gains from 4K to 1080p at any given setting are modest—for example, 50 FPS at 1080p Low versus 30 FPS at 4K Low—which points to a GPU that is working at near maximum capacity even at the lowest resolution. The measured FPS at 1080p Ultra (28 FPS) is barely higher than 4K Low (30 FPS), reinforcing that the GPU's limitations dominate across the entire resolution spectrum.

Measured FPS Breakdown

At 1920x1080, the lowest tested resolution, the combo produces an average of 50 FPS on Low settings, which is the only configurable preset that approaches playability. Medium settings drop the average to 39 FPS, with a minimum of 33 FPS and a maximum of 45 FPS. High settings reduce the average further to 30 FPS, and Ultra bottoms out at 28 FPS. None of these 1080p results reach the 60 FPS threshold that many flight sim enthusiasts consider smooth, and the gap between Low and Ultra is only 22 FPS.

Moving to 2560x1440, the performance degrades predictably. Low settings yield an average of 41 FPS, a 9 FPS reduction from 1080p Low. Medium averages 32 FPS (min 27, max 36), High averages 24 FPS, and Ultra averages 23 FPS. The 1440p results show a compression of the settings range: the spread between Low and Ultra narrows to 18 FPS, compared to 22 FPS at 1080p, because the GPU is increasingly unable to differentiate between presets as pixel count rises.

At 3840x2160, the performance falls below playable thresholds for all settings. Low averages 30 FPS, Medium averages 23 FPS (min 20, max 27), High averages 18 FPS, and Ultra averages 17 FPS. The 4K results are notable for how tightly clustered they are: the difference between Low and Ultra is just 13 FPS. This suggests that at 4K, the GPU is so heavily pixel-bound that even reducing the graphics quality settings has limited impact on frame rates. The Medium preset's minimum of 20 FPS and maximum of 27 FPS indicates significant frame time variance, which would manifest as stuttering during flight.

The data across all 12 measured rows shows a consistent pattern: the Arc B370 cannot deliver a smooth experience at any resolution/settings combination. The best-case scenario is 1080p Low at 50 FPS average, but the lack of a minimum FPS figure for that row means we cannot assess frame consistency. The only rows with min/max data are the Medium presets, which show a roughly 12-13 FPS range at each resolution—a narrow band that suggests the GPU is operating at a hard performance ceiling.

CPU Role

The AMD Ryzen 9 9900X is a 12-core, 24-thread processor based on the Zen 5 architecture (Granite Ridge), manufactured on TSMC's 4 nm process. It has a base clock of 4.40 GHz and a boost clock of 5.60 GHz, with 64 MB of L3 cache and 1 MB of L2 cache per core. The CPU supports DDR5 memory in a dual-channel configuration, with a theoretical memory bandwidth of 89.6 GB/s. It also features 24 PCIe Gen 5 lanes (CPU only) and an integrated Radeon Graphics solution, though the latter is irrelevant given the discrete-class Arc B370 IGP.

Benchmark data shows the CPU's strengths. The Cinebench R23 multicore score is 32172, and the single-core score is 2253. Geekbench results are 22174 multicore and 3010 single-core. PassMark multithread scores 54643, with a single-thread score of 4672. The CPU's average benchmark score of 57498 places it in the 92nd percentile of all CPUs. Its nearest rivals include the AMD EPYC 9015 (scoring 57555, just 0.1% higher), the Intel Core i9-14900 (scoring 58115, 1.1% higher), and the Intel Xeon Platinum 8260M (scoring 58323, 1.4% higher). The Ryzen 9 9900X actually outperforms the AMD EPYC 7313, which scores 57399 (0.2% lower).

In Microsoft Flight Simulator, the CPU's role is to handle the simulation's complex physics, flight dynamics, and AI traffic. The game is known to be single-thread-bound in certain scenarios, but the 9900X's 5.60 GHz boost clock and strong single-core performance (3dmark single-thread score of 1286, Cinebench R15 single-core of 353) suggest it would handle these workloads well if the GPU were not limiting. However, the measured FPS data shows that the CPU's capabilities are never fully exercised. At 1080p Low, where CPU load would be highest relative to GPU load, the average FPS is only 50—a figure that a much weaker CPU could achieve with a proper GPU. The 12 cores and 24 threads are essentially idle while waiting for the Arc B370 to produce frames.

The CPU's 120 W TDP and active production status indicate a high-end desktop part, and its launch MSRP was $499. But in this combination, that investment yields no benefit in game performance. The data suggests that even the integrated Radeon Graphics on the Ryzen 9 9900X might perform similarly to the Arc B370, given the latter's 5th percentile GPU ranking. The CPU's 3dmark max-thread score of 13929 and PassMark physics score of 3381 further demonstrate that it has ample headroom for simulation workloads, yet the GPU's 6.144 TFLOPS FP32 throughput is the hard ceiling.

Settings Recommendations

Given the measured data, the only viable preset for any semblance of playability is 1080p Low, which yields an average of 50 FPS. This is the single row that exceeds 30 FPS by a comfortable margin, and it is the only configuration that could be considered acceptable for casual flight. All other settings at 1080p fall below 40 FPS, with Ultra dipping to 28 FPS. At 1440p, Low averages 41 FPS, which might be playable for users who prioritize resolution over smoothness, but the 4K results are uniformly below 30 FPS and should be avoided.

The Medium presets provide the only min/max data in the measured rows. At 1080p Medium, the average is 39 FPS with a range of 33-45 FPS. At 1440p Medium, the average is 32 FPS with a range of 27-36 FPS. At 4K Medium, the average is 23 FPS with a range of 20-27 FPS. These figures show that Medium settings offer a slightly more stable frame rate than Low or High, which lack min/max data and may exhibit more variance. However, the absolute FPS values are still low for a flight simulator, where smooth camera panning is essential.

For users who must run this combination, the recommendation is to prioritize resolution over settings. The data shows that dropping from 1080p Low to 1440p Low costs 9 FPS (from 50 to 41), while dropping from 1080p Low to 1080p Medium also costs 11 FPS (from 50 to 39). The FPS penalty for moving to 1440p Low is less severe than the penalty for raising settings to Medium at 1080p, suggesting that the GPU can handle a moderate resolution increase better than it can handle increased graphical complexity. However, the 4K Low result of 30 FPS is only 20 FPS below 1080p Low, which indicates that even at the lowest settings, 4K is not a realistic target.

The Ultra preset is uniformly the worst performer, with averages of 28 FPS at 1080p, 23 FPS at 1440p, and 17 FPS at 4K. There is no scenario where Ultra provides an acceptable experience, and the gap between Ultra and High is only 2-6 FPS depending on resolution. This narrow margin suggests that the Ultra preset's additional graphical features (likely enhanced shadows, reflections, and terrain detail) are not the primary bottleneck—the GPU is simply too weak to benefit from the extra detail. Users should avoid Ultra entirely and instead use Low or Medium at 1080p, accepting that the experience will be more about function than visual fidelity.

In summary, the data supports 1080p Low as the only setting that approaches a playable frame rate, with 1440p Low as a secondary option for users who prefer higher resolution. All Medium, High, and Ultra settings should be avoided, as they yield average FPS below 39 at any resolution. The rig's CPU is vastly more capable than the GPU, and no settings adjustment can overcome the Arc B370's fundamental performance limitations in this simulation.