Fortnite

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

The Intel Core i3-12100F is a 4-core, 8-thread Alder Lake-S desktop processor with a base clock of 3.30 GHz and a boost clock of 4.30 GHz. It carries 12 MB of shared L3 cache, with 80 KB of L1 and 1.25 MB of L2 per core, and supports both DDR4 and DDR5 memory through a dual-channel bus. This is a modest CPU by modern standards, sitting at the 68th percentile of all processors in the database, with an average benchmark score of 13,494.

The data shows this processor is a clear limiting factor in Fortnite when paired with the RTX 5090. Look at the 1080p Low preset: the average FPS is 412. At 1440p Low, it drops to 311. At 4K Low, it falls to 164. The scaling pattern is revealing — from 1080p to 1440p, the drop is about 24.5%, and from 1440p to 4K, the drop is about 47.3%. If the GPU were the bottleneck across all resolutions, the Low preset scores would remain far more stable as resolution increases, since Low settings reduce GPU load substantially. Instead, the massive FPS collapse at higher resolutions with Low settings suggests the CPU is struggling to feed the GPU, and then the GPU load itself becomes prohibitive at 4K.

The CPU’s single-threaded and multi-threaded benchmark results put its capabilities into context. In Cinebench R23, it scores 1,674 single-core and 11,860 multi-core. In Geekbench, it scores 1,932 single-core and 7,252 multi-core. PassMark single-thread score is 3,444, with a multithread score of 14,015. These numbers indicate a solid budget performer, but Fortnite’s engine benefits from higher core counts and faster memory subsystems — the i3-12100F’s 4-core/8-thread configuration can deliver high frame rates at 1080p, but the ceiling is visible when frame rates approach 400 FPS.

The nearest rival data reinforces this positioning. The Intel Core 3 N355 scores 13,492, a 0% delta; the Intel Core i5-9500 scores 13,452, a 0.3% delta; the Intel Core 5 120UL scores 13,594, a -0.7% delta; and the Intel Core i7-1250U scores 13,351, a 1.1% delta. All four rivals are within roughly 1% of the i3-12100F’s average score, meaning the CPU is competitively matched with those parts in general compute, but none of them are high-end gaming processors either. This CPU is built for value-oriented systems, and in this pairing, it’s clearly the weaker link.

The fact that the combo ranks 132nd out of 3,700 tested combinations in Fortnite — placing it in roughly the 96.4th percentile of all combos — shows that even with a middle-tier CPU, the RTX 5090’s raw power lifts this system far above the vast majority of tested configurations. However, the rank also implies that many other CPU+GPU pairings achieve higher Fortnite performance, and the CPU is a primary reason this combo doesn’t rank higher.

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

The NVIDIA GeForce RTX 5090 is a flagship Blackwell 2.0 GPU built on TSMC’s 5 nm process, with 92,200 million transistors on a 750 mm² die. It features 32 GB of GDDR7 memory on a 512-bit bus, delivering 1.79 TB/s of bandwidth. The base clock is 2,017 MHz with a boost clock of 2,407 MHz, and memory runs at 1,750 MHz (28 Gbps effective). The GPU has 21,760 shading units, 680 texture mapping units, 176 ROPs, 170 ray tracing cores, and 680 tensor cores. Its compute capability is substantial: 104.8 TFLOPS FP32, 423.6 GPixel/s pixel rate, and 1,636.8 GTexel/s texture rate.

The RTX 5090 sits at the 92nd percentile of all GPUs in the database, with an average benchmark score of 79,842. Its nearest rivals are the NVIDIA Tesla P100 PCIe 16 GB (79,605, 0.3% delta), Tesla P100 PCIe 12 GB (79,396, 0.6% delta), AMD Radeon RX 6850M XT (78,940, 1.1% delta), and AMD Radeon Pro Vega 64X (80,959, -1.4% delta). These are all high-performance compute-oriented or previous-generation parts, and the RTX 5090 beats three of them while trailing the Pro Vega 64X by a small margin in average score.

In Fortnite, the GPU’s role is complex because of the CPU bottleneck. At 1080p Low, the system hits 412 FPS average — this is near the practical limit for many displays, and it’s likely the CPU is preventing even higher numbers. At 1080p Ultra, the average is 176 FPS, a 57.3% drop from Low. This indicates that at 1080p, the GPU is doing more work as settings increase, but it still has headroom — the CPU can’t push past its limits.

The 32 GB VRAM is far more than Fortnite requires at any resolution or setting, so memory capacity is not a constraint in any measured configuration. The 1.79 TB/s bandwidth and 512-bit bus ensure the GPU can feed its 21,760 shading units without memory starvation, which is critical at 4K Ultra where pixel fill requirements are extreme. The 4K Ultra result of 59 FPS average shows that even this flagship GPU is challenged by the most demanding settings at maximum resolution — but it still delivers a playable frame rate.

The GPU’s benchmark scores in DirectX 12 (PassMark DX12 score of 185) and DirectX 11 (341) are lower than expected relative to its raw compute scores, which suggests driver or API overhead considerations. However, the 3DMark Steel Nomad DX12 score of 18,355 and Geekbench Vulkan score of 376,728 indicate strong modern API performance. In Fortnite, the data shows the GPU is the dominant factor only at 4K Ultra, where it’s fully saturated; everywhere else, CPU limitations cap performance.

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

The measured FPS data across resolutions tells a clear story about which component is the bottleneck at each setting tier. Starting with the Low preset, which minimizes GPU load: 1080p averages 412 FPS, 1440p drops to 311 FPS (a 24.5% decrease), and 4K drops to 164 FPS (a 47.3% decrease from 1440p, 60.2% from 1080p). This pattern — a massive collapse at 4K — indicates that even at Low settings, the GPU becomes overwhelmed at 3840x2160 because there are simply too many pixels to render, despite the CPU’s limitations.

For the Medium preset, which includes measured minimum and maximum FPS at all resolutions, the scaling is: 1080p averages 345 FPS (min 293, max 396), 1440p averages 219 FPS (min 186, max 252), and 4K averages 116 FPS (min 98, max 133). The drop from 1080p to 1440p is 36.5%, and from 1440p to 4K is 47.0%. The larger drop at higher resolutions compared to the Low preset suggests that as settings increase, the GPU becomes more of a limiting factor, but the CPU still constrains the 1080p result — 345 FPS on Medium is close to the Low preset’s 412 FPS, showing that the CPU is near its ceiling around 400 FPS.

The High preset shows: 1080p at 267 FPS, 1440p at 170 FPS, and 4K at 90 FPS. The percentage drops are 36.3% (1080p to 1440p) and 47.1% (1440p to 4K). These are nearly identical to the Medium preset’s scaling, which suggests a consistent pattern: once the GPU is loaded sufficiently, the scaling penalty is roughly 36-47% per resolution step. The Ultra preset, however, shows a different story: 1080p at 176 FPS, 1440p at 112 FPS, and 4K at 59 FPS. The drops are 36.4% and 47.3%, respectively — the same proportional pattern, but starting from a lower absolute baseline because Ultra settings are far more demanding.

The key insight from the resolution scaling is that the CPU is the primary limiter at 1080p across all settings. At 1080p Low (412 FPS) versus 1080p Ultra (176 FPS), the GPU is doing far more work in the latter, yet the CPU still delivers 176 FPS — a playable number. At 4K, the GPU becomes the dominant bottleneck, particularly at Ultra where it manages just 59 FPS. The fact that the scaling percentages are so consistent (roughly 36-37% for the first step and 47% for the second step) across Medium, High, and Ultra suggests that both components are contributing to the limits in a predictable way, with the GPU’s pixel throughput being the main constraint at 4K.

FAQ

Q: Is the Intel Core i3-12100F a bottleneck for the RTX 5090 in Fortnite?

A: Yes, the data indicates the CPU limits performance at 1080p and 1440p. At 1080p Low, the system hits 412 FPS, but at 4K Low it drops to 164 FPS — a 60.2% decrease. The CPU’s average benchmark score of 13,494 places it at the 68th percentile, while the GPU is at the 92nd percentile, showing a significant capability gap that manifests as CPU-bound performance at lower resolutions.

Q: How does the RTX 5090 handle 4K Ultra in Fortnite?

A: At 4K Ultra, the system averages 59 FPS. This is playable but not high-refresh territory. The GPU’s 32 GB of GDDR7 memory and 1.79 TB/s bandwidth are sufficient for the game’s demands, but the 4K Ultra workload saturates even this flagship GPU, which has 21,760 shading units and 176 ROPs.

Q: What settings give the highest frame rates?

A: The Low preset at 1080p delivers the highest measured average at 412 FPS. At 1440p, Low gives 311 FPS, and at 4K, Low gives 164 FPS. The Medium preset at 1080p yields 345 FPS, which is 67 FPS lower than Low but still very high.

Q: How does this combo rank against other tested configurations?

A: This combo ranks 132nd out of 3,700 tested combinations in Fortnite, placing it in roughly the 96.4th percentile. This is an excellent rank, driven primarily by the GPU’s raw power, though the CPU prevents it from ranking higher.

Q: Is the 32 GB VRAM useful for Fortnite?

A: Fortnite does not require anywhere near 32 GB of VRAM at any measured resolution or setting. The memory capacity is not a limiting factor in any configuration tested. The GPU’s bandwidth and compute are the relevant performance attributes for this game.

Q: What is the performance difference between the Low and Ultra presets at each resolution?

A: At 1080p, Low averages 412 FPS versus Ultra’s 176 FPS, a 57.3% difference. At 1440p, Low averages 311 FPS versus Ultra’s 112 FPS, a 64.0% difference. At 4K, Low averages 164 FPS versus Ultra’s 59 FPS, a 64.0% difference. The gap widens at higher resolutions because the GPU becomes more saturated.

Settings Recommendations

Based on the measured FPS rows, the optimal settings depend entirely on the target resolution and the user’s refresh rate. The data shows a clear trade-off: higher settings reduce FPS significantly, but the game remains playable even at the most demanding configuration.

For 1080p displays, the Medium preset at 345 FPS average (min 293, max 396) provides an excellent experience with a large safety margin above 240 Hz refresh rates. The High preset at 267 FPS is also viable for 240 Hz displays, while Ultra at 176 FPS is suitable for 165 Hz monitors. The Low preset at 412 FPS is excessive for any current display and sacrifices visual quality for numbers that exceed what monitors can show.

For 1440p, the Medium preset at 219 FPS (min 186, max 252) is the best balance — it clears 144 Hz comfortably and even approaches 240 Hz territory in the maximum FPS reading. The High preset at 170 FPS is excellent for 144 Hz displays, while Ultra at 112 FPS is adequate for 100 Hz panels but not for higher refresh rates. Low at 311 FPS is again beyond what typical 1440p monitors can display.

For 4K, the situation is more constrained. The Medium preset at 116 FPS (min 98, max 133) is the recommended choice for 120 Hz displays, as it stays above 98 FPS even in the minimum reading. High at 90 FPS is suitable for 90 Hz panels, while Ultra at 59 FPS is only appropriate for 60 Hz displays. Low at 164 FPS on a 4K panel is possible but would require a high-end 4K high-refresh monitor to be useful.

The data suggests that Medium settings at 1080p and 1440p, and Medium at 4K, offer the best experience because they provide high average FPS with acceptable minimum frame rates. The Medium preset’s minimum FPS readings (293 at 1080p, 186 at 1440p, 98 at 4K) indicate stable performance without severe drops, which is more important for gameplay smoothness than raw average FPS. Ultra settings provide a 57-64% FPS penalty compared to Low, which is a steep cost for visual improvements that may not be noticeable in fast-paced battle royale gameplay.

Measured FPS Breakdown

The measured FPS data provides exact figures for 12 configurations (3 resolutions × 4 settings presets). This section details every measured row from the FACT PACK.

1920x1080 (1080p):

• Low: 412 FPS average. This is the highest measured FPS in the entire dataset, and it’s likely CPU-limited — the i3-12100F’s 4-core/8-thread design can only push so many frames, even with the RTX 5090’s massive GPU headroom at Low settings.
• Medium: 345 FPS average, with a minimum of 293 FPS and maximum of 396 FPS. The 52 FPS gap between minimum and maximum indicates some variability, but the minimum is still far above any common refresh rate.
• High: 267 FPS average. This is a 22.6% drop from Medium, showing how the GPU load increases with settings even at 1080p.
• Ultra: 176 FPS average. This is a 57.3% drop from Low and a 34.1% drop from High, demonstrating that Ultra settings impose a substantial GPU workload even at the lowest resolution.

2560x1440 (1440p):

• Low: 311 FPS average. This is 24.5% lower than the 1080p Low result, showing that the resolution increase alone has a significant impact.
• Medium: 219 FPS average, with a minimum of 186 FPS and maximum of 252 FPS. The minimum is still above 144 Hz, making this a strong configuration for high-refresh 1440p monitors.
• High: 170 FPS average. This is a 22.4% drop from Medium, and it remains above 144 Hz — a good match for common 1440p gaming displays.
• Ultra: 112 FPS average. This is a 64.0% drop from Low and a 34.1% drop from High, mirroring the proportional penalties seen at 1080p.

3840x2160 (4K):

• Low: 164 FPS average. This is a 47.3% drop from the 1440p Low result, and a 60.2% drop from 1080p Low. The GPU is now the primary bottleneck even at Low settings.
• Medium: 116 FPS average, with a minimum of 98 FPS and maximum of 133 FPS. The minimum of 98 FPS is just below 100 Hz, which is a meaningful threshold for many high-refresh monitors.
• High: 90 FPS average. This is a 22.4% drop from Medium, consistent with the scaling pattern observed at lower resolutions.
• Ultra: 59 FPS average. This is the most demanding configuration in the dataset, and it’s the only measured result below 60 FPS. It represents a 64.0% drop from Low and a 34.4% drop from High.

The overall pattern is consistent: Medium and High presets scale similarly (roughly 22% penalty from Medium to High at every resolution), while Ultra imposes a much larger penalty (roughly 34% from High to Ultra). The resolution scaling from 1080p to 1440p costs roughly 36-37% of FPS, and from 1440p to 4K costs roughly 47%, regardless of settings. This consistency suggests that both the CPU and GPU contribute to performance limits in predictable ways, with the CPU capping 1080p results and the GPU becoming the dominant constraint at 4K.