Palworld

Palworld on the Intel Core Ultra 7 265 and Intel Arc B390 combination delivers a playable experience only at the most forgiving settings, with the data revealing a hard ceiling imposed by the graphics solution. The measured frame rates across all resolutions and presets show a system that is consistently GPU-bound, struggling to maintain fluid motion even when the CPU has ample headroom. This analysis breaks down the measured FPS, scaling behavior, and the respective roles of the processor and integrated graphics in shaping the final performance picture.

Resolution Scaling

The performance drop from 1080p to 4K is steep and consistent, underscoring that the Intel Arc B390 is the primary bottleneck. At Low settings, the average frame rate falls from 84 FPS at 1920x1080 to 56 FPS at 2560x1440, a reduction of 28 FPS, and then further down to 31 FPS at 3840x2160. This represents a 63% performance loss from 1080p to 4K, a clear indicator that the rendering pipeline is saturated by the increasing pixel count. The scaling is not linear; the jump from 1440p to 4K costs 25 FPS, while the jump from 1080p to 1440p costs 28 FPS, suggesting the GPU's memory bandwidth and pixel throughput are being exhausted at the higher resolutions.

The pattern repeats across all presets, though the absolute deltas shrink as the settings become more demanding. At High settings, the average FPS drops from 53 at 1080p to 35 at 1440p and finally to 19 at 4K. The difference between 1080p and 4K is 34 FPS, a 64% decline, nearly identical to the Low preset's relative loss. This consistency indicates that resolution is the dominant factor, not the complexity of the scene or the specific graphical features enabled. The frame rate is being throttled by the fill rate and shading capacity of the Arc B390, which is a mobile-class integrated part, regardless of the preset selected.

At Ultra settings, the situation becomes nearly unplayable at higher resolutions. The average FPS at 1080p is 40, dropping to 26 at 1440p and a mere 15 at 4K. The scaling from 1440p to 4K shows a 11 FPS drop, which is proportionally similar to other presets, but the starting point is so low that the final result is a slideshow. The data suggests that the limiting component is unequivocally the GPU; the CPU's performance is not a factor in these scaling trends because the frame rate is dictated by how quickly the integrated graphics can process and output frames, not by how quickly the CPU can feed it data. The steep decline in FPS with each resolution step is the classic signature of a pixel-bound workload.

Settings Recommendations

The measured data shows that the only viable path to a smooth experience is to prioritize the Low preset and accept lower resolutions. At 1920x1080 with Low settings, the combo achieves an average of 84 FPS, which is the only configuration in the entire dataset that approaches high-refresh-rate territory. This is the clear best choice for playability. The Medium preset at 1080p delivers 68 FPS, which is still playable, but the 16 FPS penalty compared to Low is significant for what is likely a modest visual upgrade. The step from Medium to High costs another 15 FPS, landing at 53 FPS, which is borderline for a survival game that benefits from smooth camera movement.

For users with 1440p displays, the choices are more constrained. Low settings at 2560x1440 yields an average of 56 FPS, which is acceptable for a slower-paced game, but Medium drops to 45 FPS and High to 35 FPS. The data indicates that High and Ultra at 1440p are not recommended, as the frame rates fall below the threshold for comfortable play. At 4K, the situation is dire; even Low settings only manage an average of 31 FPS, with Ultra bottoming out at 15 FPS. The recommendation based on the measured rows is to use Low settings for any resolution above 1080p, and to consider 1080p Medium as a compromise if visual fidelity is prioritized over raw frame rate.

The frame time consistency, where available, also informs the recommendation. The Medium preset at 1080p shows a minimum of 58 FPS and a maximum of 79 FPS, a range of 21 FPS. At 1440p Medium, the range is from 38 to 52 FPS. These ranges are relatively tight, indicating that the frame pacing is stable even when the average is low. This suggests that the Low preset at 1080p, with its higher average, is likely to provide a consistent and stutter-free experience. Given the data, the best experience per the measured rows is unequivocally 1080p Low, with 1080p Medium as a secondary option for users who prefer higher visual quality.

How This Combo Ranks

The Intel Core Ultra 7 265 and Intel Arc B390 combination ranks 2607 out of 2944 tested combos in Palworld. This places it in the bottom 11.5% of all system configurations, a low position that reflects the integrated GPU's limitations. The rank is a direct consequence of the Arc B390's performance class; the CPU is a high-end desktop part, but the GPU is designed for power efficiency, not gaming throughput. The data shows that this pairing is heavily imbalanced for gaming purposes, with the GPU unable to leverage the CPU's capabilities.

The GPU's individual percentile ranking reinforces this assessment. The Intel Arc B390 sits in the 9th percentile of all GPUs, with an average benchmark score of 1482 from the 3DMark Steel Nomad test. Its nearest rivals are low-end dedicated parts from a previous era, such as the NVIDIA GeForce GT 520MX, which scores 1463 (a 1.3% difference), and the NVIDIA GeForce GT 710, which scores 1443 (a 2.7% difference). This places the Arc B390 in the same performance tier as entry-level discrete graphics cards from over a decade ago, confirming that its gaming capability is minimal.

The CPU, in contrast, performs exceptionally well. The Intel Core Ultra 7 265 is in the 93rd percentile of all CPUs, with an average benchmark score of 64640. Its nearest rivals include the AMD EPYC 4464P, which scores 64823 (a -0.3% difference, meaning the Intel part is slightly ahead), and the Intel Core Ultra 7 265F, which scores 64438 (a 0.3% difference). This indicates that the processor is performing exactly as expected for its class, offering top-tier compute performance that is completely wasted in this gaming scenario. The combination's low rank is solely attributable to the GPU, which cannot translate the CPU's power into frame rates.

Measured FPS Breakdown

The dataset provides a complete matrix of average frame rates across four resolutions and four settings presets. At 1920x1080, the results are as follows: Low averages 84 FPS, Medium averages 68 FPS (with a minimum of 58 and maximum of 79), High averages 53 FPS, and Ultra averages 40 FPS. Moving to 2560x1440, the averages drop to 56 FPS for Low, 45 FPS for Medium (minimum 38, maximum 52), 35 FPS for High, and 26 FPS for Ultra. At 3840x2160, the averages fall to 31 FPS for Low, 25 FPS for Medium (minimum 21, maximum 29), 19 FPS for High, and 15 FPS for Ultra.

The data shows a consistent pattern: each step up in settings costs a significant amount of performance. From Low to Medium at 1080p, there is a 16 FPS drop. From Medium to High, a further 15 FPS drop. From High to Ultra, an additional 13 FPS drop. This pattern is similar at 1440p, with drops of 11 FPS, 10 FPS, and 9 FPS for the same respective steps. At 4K, the drops are 6 FPS, 6 FPS, and 4 FPS. The absolute delta between presets shrinks as resolution increases because the GPU is already overwhelmed, and the additional load from higher settings has a diminishing absolute impact.

The minimum and maximum FPS figures for the Medium preset provide insight into frame time consistency. At 1080p, the range is 58 to 79 FPS, indicating a 21 FPS variance. At 1440p, the range is 38 to 52 FPS, a 14 FPS variance. At 4K, the range is 21 to 29 FPS, an 8 FPS variance. The variance decreases with resolution, but this is not a sign of stability; it is a sign that the GPU is pegged at its maximum capacity, and the frame rate is tightly clustered around a low average. The lack of min/max data for other presets means that the Medium preset is the only one where the data suggests a stable output, but the absolute numbers are too low for a good experience at higher resolutions.

GPU Role

The Intel Arc B390 is an integrated graphics processor based on the Xe3-LPG architecture, built on a 3 nm process. It operates with a base clock of 300 MHz and a boost clock of 2500 MHz. The memory is System Shared, meaning it uses a portion of the main system RAM, with bandwidth listed as System Dependent. This is a critical limitation, as the GPU competes with the CPU for memory bandwidth, which is a primary reason for the poor scaling at higher resolutions. The GPU has 1536 shading units, 48 TMUs, 24 ROPs, and 12 ray tracing cores, providing a theoretical FP32 performance of 7.680 TFLOPS.

The 3DMark Steel Nomad score of 1482 places the Arc B390 in the 9th percentile of all GPUs. Its nearest rivals are the NVIDIA GeForce GT 520MX (score 1463, 1.3% faster), the NVIDIA GeForce 800M (score 1460, 1.5% faster), and the NVIDIA GeForce GT 625 OEM (score 1446, 2.5% faster). These are all low-end parts, and the fact that the Arc B390 is comparable to them indicates that its gaming performance is severely limited. The pixel rate of 60.00 GPixel/s and texture rate of 120.0 GTexel/s are theoretical maxima that are unlikely to be achieved in a real-world game like Palworld due to the shared memory bottleneck.

The GPU's specifications directly explain the measured frame rates. The System Shared memory means that the effective bandwidth is dependent on the system's DDR5 RAM configuration, which is shared with the CPU. In Palworld, a game with large open-world environments and many dynamic objects, the constant data streaming from memory to the GPU is a bottleneck. The low ROP count of 24 also limits the fill rate, which is why the frame rate drops so precipitously at 4K. The GPU is simply not designed for the demands of modern gaming at high resolutions, and the data confirms that it is the limiting factor in this combination.

CPU Role

The Intel Core Ultra 7 265 is a 20-core, 20-thread desktop processor based on the Arrow Lake architecture, built on a 3 nm process. It has a base clock of 2.40 GHz and a boost clock of 5.30 GHz, with a TDP of 65 W. The cache hierarchy includes 192 KB of L1 per core, 3 MB of L2 per core, and 30 MB of shared L3 cache. It supports dual-channel DDR5 memory with a bandwidth of 102.4 GB/s. The processor is in the 93rd percentile of all CPUs, with an average benchmark score of 64640, placing it in the upper echelon of desktop processors.

In synthetic benchmarks, the CPU shows strong results. It scores 42216 in Cinebench R23 multi-core and 5960 in single-core. The PassMark multi-thread score is 49682, with a single-thread score of 4689. These figures indicate a very capable processor for compute-heavy tasks. Its nearest rivals are the AMD EPYC 4464P (score 64823, -0.3% delta), the Intel Core Ultra 7 265F (score 64438, 0.3% delta), and the AMD EPYC 7343 (score 64202, 0.7% delta). This places it in a tightly contested group of high-end server and desktop parts, confirming its strong multi-threaded capabilities.

In the context of Palworld, however, the CPU's power is irrelevant to the final frame rate. The measured FPS is entirely dictated by the GPU. The data shows that even at 1080p Low, where the CPU has the least amount of work to do relative to the GPU, the frame rate is only 84 FPS. This is far below what the CPU is capable of supporting; a processor of this class can typically push frame rates well over 100 FPS in most games. The 20 cores and 20 threads are underutilized, as the game is not heavily threaded, and the high boost clock of 5.30 GHz is not a factor because the GPU cannot render frames fast enough to require more CPU input. The CPU is a passenger in this system, waiting for the GPU to catch up.