Inside

Inside, the puzzle platformer released on July 6, 2016, shows remarkably high frame rates with the Intel Core i5-13400F paired with the NVIDIA GeForce RTX 4090. Measured data records average FPS from 140 at 4K Ultra to 333 at 1080p Low, indicating that this combination decisively outclasses the game’s demands. The rankings place this combo at 321 of 3,716 tested, which is a strong position given the breadth of tested combinations. The following sections detail settings, GPU contribution, resolution scaling, and measured frame data.

Settings Recommendations

The data set provides avgFps and, for the Medium preset only, min and max values at every resolution. When choosing a preset, the presence of full timing data is a practical signal: the Medium preset has a recorded minimum at each resolution, binding the floor of what the hardware delivers accurately. Without a recorded min or max for Low, High, or Ultra, only the average is known for those presets.

At 3840x2160, the Ultra preset returns an average of 140 FPS, High averages 147, Medium captures an avgFps of 235 with a min of 200 and max of 270, and Low 255. The difference between High and Medium at 4K is roughly 88 FPS, which is significant, but note that High lacks min/max data. For players who want the highest absolute frame rate, Low at 255 is the fast preset, but the Medium at 235 plus known minimum 200 is the highest preset that includes an explicit frame floor. Therefore, on a 4K display, Medium appears the best measured compromise: it stays above 235 FPS on average, and the recorded low is 200, which is still far above the game’s likely animation frame requirement.

At 2560x1440, Medium delivers an average of 275 FPS with min 234 and max 317. Low is 298, High 229, Ultra 222. Here the gap from Medium to Low is only 23 FPS on the avg, so the preselect with the higher visual content is nearly as fast. High and Ultra drop below 230 FPS, which is still high. At this resolution, Medium is again the stronger pick due to the min/max data, while Ultra still averages 222 FPS, which remains a plausible setting if visual detail is a high priority.

At 1920x1080, the values are 333 (Low), 307 (Medium), 256 (High), 252 (Ultra). The Medium has a min of 261 and a max of 353, meaning it shows the least stable frame in the set; but even a foot at 261 FPS is well above any smoothness threshold. The difference between High and Ultra is a mere * 4 FPS but the pack does not record a minimum or max for those. So if you want to keep the highest visual detail without missing a known floor, the Medium preset again gives the only measured lower bound at 261 FPS. No preset is a bad choice; even the Ultra average at 1080 exceeds 240 FPS.

In short, default to Medium for a stable baseline with confirmed min and max data. At the highest visual level, Ultra is still playable, but for reproducible behaviour and predictable latencies, Medium is the sweet spot in the measured dataset.

GPU Role

The RTX 4090 is the central element in this test. It uses the Ada Lovelace architecture with an AD102 chip on a 5 nm TSMC process. The GPU memory is 24 GB of GDDR6X on a 384-bit bus, providing a bandwidth of 1.01 TB/s. The base clock is 2235 MHz and boost 2520 MHz. The card is honestly triple-slot, and its TDP is 450 W. These figures are from its specification sheet, and while we cannot directly measure power draw in this page, those numbers frame the performance capability.

In the measured FPS, the GPU is clearly not a bottleneck at 1080p, given the average exceeds 252 FPS in Ultra. At 4K, however, the load is heavier. For example, 4Kp Ultra gives only 140 average, while at 1080p Ultra gives 252, a difference of 112 FPS. This suggests that the 4090’s 4K output is limited by the modern fill rate and memory throughput, even though its bandwidth is 1.01 TB/s and its 82.58 TFLOPS is the second highest among consumer cards (mechanically). The GPU’s pixel rate is 443.5 GPixel/s and texture rate 1,290.2 GTexel/s, which we can mention as context.

In terms of raw benchmark numbers, the RTX 4090 has an average benchmark score of 60,347. Its nearest rival in this database is the Intel Arc Pro A60, which has a similar average score of ?? Actually, the nearestRivals list does not contain an Arc Pro A60? Wait, it has "name": "AMD Radeon Pro V710"? Let's look at the GPU nearestRivals: they are:

• AMD Radeon Pro V710 (avgScore 58657, deltaPct 2.9)
• AMD Radeon Pro W6600M (avgScore 61896, deltaPct -2.5)
• AMD Radeon Pro Vega 48 (avgScore 60140, deltaPct 0.3)
• Intel Arc Pro A60 (avgScore 60326, deltaPct 0) — actually the pack says "Intel Arc Pro A60". So we write that.

We can state: "The GPU’s average benchmark index is within 0.3% of the AMD Radeon Pro Vega 48 and essentially matches the Intel Arc Pro A60’s 60,326." But we must not write percentages; we can write "The GPU's average benchmark score is 60347, compared to the Intel Arc Pro A60 for which deltaPct is 0, meaning the average scores are equivalent; the AMD Radeon Pro Vega 48 has a deltaPct of 0.3, meaning it is 0.3 points higher" — but 0.3 is a number, write it. However, we cannot write "0.3%", we can write "deltaPct of 0.3" – that is a field. We'll say "the deltaPct for the AMD Radeon Pro Vega 48 is 0.3, which means the benchmark score is 0.3 times higher? It's a percentage difference. The rules says "no percentages" unless exact number is in the pack. Indeed, the number 0.3 appears in the pack as deltaPct. So we can mention it as a value.

We can focus on the GPU role in this title by noting that the 24 GB buffer is enough for any texture or shader output in this game, but we have no direct evidence. No better to say the "24GB buffer, a 1.01 TB/s bandwidth, and the boost clock of 2520 MHz help support the framerate numbers." But we can't claim "support" without proof. We'll just list specs.

The real GPU effect on the frame is visible in the resolution scaling: the FPS drops as resolution increases, which is typical.

Resolution Scaling

Scaling at all presets follows the expected slope: the frame rate decreases when moving to a higher resolution. For the High preset, the measured average FPS is 256 at 1920x1080, 229 at 2560x1440, and 147 at 3840x2160. For Low, it is 333 at 1080p, 298 at 1440p, and 255 at 4K. This reduction is consistent across Medium and Ultra as well: Medium goes from 307 at 1080p to 275 at 1440p to 235 at 4K; Ultra goes 252, 222, and 140. The decrease from 1080p to 4K is substantial: for High it is about 107 frames lower, for Medium about 72, for Ultra about 110, and for Low about 78. However, we cannot reduce to a percentage, we can only state the numbers.

What this scaling implies is that the GPU is the more limiting component at 4K. At 1080p, the differences between presets are narrower (worst case a 81 FPS spread from 333 to 252), which suggests that the CPU (i5-13400F) is delivering enough to keep the GPU busy but at 4K the GPU’s rendering load brings the average down. The fact that at 4K the avg FPS for High is 147, which is less than half of the 1080p of 256, shows the GPU is the dominant resource.

The only preset with min/max data, Medium, shows: 1080p: 261 to 353, 1440p: 234 to 317, 4K: 200 to 270. These numbers illustrate the frame does not drop as low as the avg would imply; the Minimum values are all well above 200 FPS, giving a stable experience.

Measured FPS Breakdown

All measurements are for the Intel Core i5-13400F paired with the RTX 4090. Below is the complete set of average (and for Medium, minimum and maximum) frames per second per settings and resolution.

| Resolution | Settings | Avg FPS | Min FPS | Max FPS |

|------------|----------|---------|---------|---------|

| 1920x1080 | Low | 333 | - | - |

| 1920x1080 | Medium | 307 | 261 | 353 |

| 1920x1080 | High | 256 | - | - |

| 1920x1080 | Ultra | 252 | - | - |

| 2560x1440 | Low | 298 | - | - |

| 2560x1440 | Medium | 275 | 234 | 317 |

| 2560x1440 | High | 229 | - | - |

| 2560x1440 | Ultra | 222 | - | - |

| 3840x2160 | Low | 255 | - | - |

| 3840x2160 | Medium | 235 | 200 | 270 |

| 3840x2160 | High | 147 | - | - |

| 3840x2160 | Ultra | 140 | - | - |

The highest average is 333 FPS at Low, 1080p. The lowest is 140 at Ultra, 4K. At 4K, the Medium preset has a floor of 200 FPS, but High and Ultra lack that information. Select minor differences: the jump from 1080 to 4K on High is 256 to 147, a 111 FPS drop. For favor the low preset, 333 diminishes to 255. The trend is clear across every column: resolution trasding average.

How This Combo Ranks

This specific combination has a measured rank of 321 out of 3716 record combos for Inside. That places it in the higher portion of the list, though not in the absolute top. The GPU has a percentile rank of 88 among all GPUs, while the CPU’s percentileOf is 77. These are the specific numbers for the hardware individually.

When compared to nearest rivals in the database: the CPU’s average benchmark score is 25292. The Intel Core 5 120 has a deltaPct of -0.3, and the AMD Ryzen 5 5600X3D also -0.3. This CPU is within measurement noise, with the AMD Ryzen 9 6900HS being exactly close deltaPct of 0. On the GPU side, the RTX 4090’s average score is 60347, and the AMD Radeon Pro Vega 48 has a deltaPct of 0.3, a near tie. The rank of 321 is a strong result; the differentiator in Inside appears to be the massive GPU.

FAQ

Q: What is the exact average FPS at 1920x1080 Ultra setting?

A: The measured average. It is 252 frames per second.

Q: Which preset includes both measured minimum and maximum FPS?

A: The Medium preset at all three resolutions: 1080p (261–353), 1440