Downfall

# Downfall Performance Analysis: Intel Core Ultra 7 265 + NVIDIA GeForce RTX 5090

The measured frame rates for this pairing in Downfall reveal a system with exceptional headroom across all resolution tiers, though the scaling patterns expose where the balance between CPU and GPU shifts. With the RTX 5090 delivering 32 GB of GDDR7 memory and the Core Ultra 7 265 providing 20 cores across Arrow Lake architecture, this combination sits at the 56th spot out of 2,928 tested combos — firmly in the top 2% of configurations. The data shows a horror title that responds aggressively to both settings changes and resolution scaling, making it a valuable case study for understanding how this flagship GPU behaves when paired with a high-core-count desktop processor.

Settings Recommendations

The measured rows across four quality presets at three resolutions paint a clear picture of how Downfall scales with graphical fidelity. At 1920x1080, the Low preset achieves 452 FPS average, Medium reaches 370 FPS, High drops to 303 FPS, and Ultra settles at 190 FPS. This progression shows that each step up in settings costs roughly 20-25% of the previous tier's performance, with the gap between Medium and High (67 FPS) being slightly larger than the gap between High and Ultra (113 FPS), suggesting diminishing returns at the highest preset.

At 2560x1440, the pattern shifts noticeably. Low delivers 323 FPS, Medium 264 FPS, High 217 FPS, and Ultra 135 FPS. The Ultra preset now takes a proportionally larger hit compared to 1080p, dropping from 190 to 135 FPS — a 29% reduction that indicates the GPU is beginning to feel the resolution pressure. Meanwhile, the Medium preset at 1440p (264 FPS) stays comfortably above the 144 Hz refresh rate threshold that many high-end monitors target, making it the sweet spot for competitive play.

The 4K results tell the most important story for settings selection. At 3840x2160, Low averages 201 FPS, Medium 164 FPS, High 135 FPS, and Ultra drops to 84 FPS. The Ultra preset at 4K falls below the 90 FPS mark, which for a horror game with likely atmospheric lighting effects could introduce noticeable judder in fast camera movements. The High preset at 135 FPS represents the best balance for 4K users — it maintains fluid motion while delivering substantially better visual quality than Medium's 164 FPS would suggest, given the relatively modest 29 FPS difference between the two.

For the best experience per the measured data, Medium at 1080p and 1440p offers the highest frame rates while still providing meaningful visual uplift over Low. At 4K, High is the recommended preset, as it keeps the average above 120 FPS while avoiding the steep performance cliff that Ultra introduces. The 140 FPS minimum recorded at Medium 4K (versus no minimum data for High) hints that Medium might offer more consistent frame pacing, but the 135 FPS average at High remains well within acceptable bounds for the genre.

GPU Role

The RTX 5090's 32 GB of GDDR7 memory across a 512-bit bus delivers 1.79 TB/s of bandwidth, which the Downfall data suggests is more than sufficient even at 4K Ultra. The fact that Ultra at 4K still manages 84 FPS average — rather than collapsing under memory pressure — indicates that the 32 GB frame buffer is not a bottleneck in this title. The GPU's 2,407 MHz boost clock and 104.8 TFLOPS FP32 throughput provide the raw compute headroom that drives the 452 FPS result at 1080p Low.

The memory clock of 1750 MHz (28 Gbps effective) paired with the 512-bit interface explains why resolution scaling remains relatively graceful. Between 1440p High (217 FPS) and 4K High (135 FPS), the drop is 82 FPS — a 38% reduction that tracks closely with the pixel count increase from 3.7 million to 8.3 million pixels. This proportionality suggests the GPU's memory subsystem is keeping pace with the rendering demands, rather than becoming a limiting factor at higher resolutions.

The RTX 5090's 170 RT cores and 680 tensor cores become relevant when considering Downfall's horror genre, which likely employs ray-traced shadows or reflections for atmospheric effect. The measured data doesn't isolate RT performance, but the overall frame rates — even at 4K Ultra with 84 FPS — imply that any ray tracing workload is being handled without catastrophic performance degradation. The GPU's 92,200 million transistors on a 5 nm TSMC process give it the architectural resources to maintain these high averages across all tested configurations.

Compared to its nearest rivals in the benchmark database, the RTX 5090's average score of 79,842 sits just 0.3% above the NVIDIA Tesla P100 PCIe 16 GB and 1.1% above the AMD Radeon RX 6850M XT. While these synthetic benchmarks don't directly translate to Downfall performance, they confirm the 5090's position as a top-tier GPU — a status the measured FPS data reinforces with triple-digit averages even at 4K.

Resolution Scaling

The transition from 1080p to 4K reveals where this system's bottleneck shifts. At Low settings, 1080p delivers 452 FPS, 1440p drops to 323 FPS (71.5% of 1080p), and 4K falls to 201 FPS (44.5% of 1080p). This scaling is almost perfectly linear with pixel count, which indicates the GPU is the primary limiting factor even at the lowest quality preset. A CPU-bound scenario would show a much flatter curve, with 1080p and 1440p delivering nearly identical frame rates.

The Medium preset tells a similar story: 370 FPS at 1080p, 264 FPS at 1440p (71.4%), and 164 FPS at 4K (44.3%). The consistency of these ratios — roughly 71% and 44% at each resolution step regardless of settings — strongly suggests that the RTX 5090's rendering throughput is the dominant constraint across all quality levels. The Core Ultra 7 265 with its 5.30 GHz boost clock is feeding the GPU fast enough that it never becomes the limiting factor.

At High settings, the pattern holds: 303 FPS at 1080p, 217 FPS at 1440p (71.6%), and 135 FPS at 4K (44.6%). Ultra follows suit with 190, 135, and 84 FPS respectively, maintaining the same approximate ratios. This uniform scaling across settings is notable because it implies that Downfall's graphical options don't disproportionately stress one component over another — each preset increases GPU workload evenly, and the CPU's 20 cores with 30 MB of L3 cache absorb the additional draw call overhead without issue.

The practical implication is that users seeking higher frame rates should prioritize resolution reduction over settings reduction. Dropping from 4K to 1440p at Ultra yields 135 FPS (a 51 FPS gain), while dropping from Ultra to High at 4K yields only a 51 FPS gain as well — but the former provides a more significant visual change. The data suggests that for maximum smoothness, 1440p Medium at 264 FPS or 1080p High at 303 FPS are the optimal configurations.

How This Combo Ranks

This specific pairing of the Intel Core Ultra 7 265 with the NVIDIA GeForce RTX 5090 achieves a combo rank of 56 out of 2,928 tested combinations in Downfall, placing it in the 98.1st percentile of all configurations. This ranking reflects the synergy between a top-tier GPU and a high-core-count CPU that, while not the absolute fastest in synthetic benchmarks, provides ample headroom for GPU-bound gaming workloads.

The CPU's percentile vs all processors stands at 93, with an average benchmark score of 64,640. Its nearest rival, the AMD EPYC 4464P, scores 64,823 — a negligible 0.3% difference. This near-parity in synthetic tests suggests that the Core Ultra 7 265 is not leaving performance on the table compared to its closest competitors. However, the EPYC processors in the rival list are server-class parts, which typically have lower clock speeds but higher core counts — the fact that they match the Ultra 7 265's average score indicates the desktop CPU's 5.30 GHz boost clock is compensating for its relatively modest 20-thread count.

The GPU's percentile of 92 and average benchmark score of 79,842 place it near the top of the database, with its closest rival being the Tesla P100 at 79,605 (0.3% lower). This 0.3% delta is within measurement noise, suggesting that the RTX 5090 and Tesla P100 perform nearly identically in aggregate benchmarks — despite the 5090's gaming-focused architecture with 21,760 shading units versus the Tesla's compute-oriented design.

That this combo ranks 56th overall — rather than top 10 — indicates that Downfall may favor slightly different hardware pairings. The gap between this configuration and higher-ranked combos likely stems from CPU choice rather than GPU, as the Core Ultra 7 265's 20 cores at 2.40 GHz base clock may not be optimal for a title that appears to scale well with fewer, faster cores based on the resolution scaling analysis. The data suggests that a CPU with higher single-thread performance could push this combo higher in the rankings.

CPU Role

The Intel Core Ultra 7 265 contributes 20 cores and 20 threads across Arrow Lake architecture, running at a base clock of 2.40 GHz and boosting to 5.30 GHz. The lack of hyperthreading (20 threads = 20 cores) means each core handles one thread, which in gaming scenarios typically reduces scheduling overhead but can limit multi-threaded throughput. The 30 MB shared L3 cache and 3 MB L2 cache per core provide substantial on-die storage for game assets, while the 102.4 GB/s dual-channel DDR5 memory bandwidth feeds both CPU and GPU data requests.

In the measured Downfall results, the CPU's role appears secondary to the GPU, but the 1080p Low result of 452 FPS reveals its capability ceiling. At this resolution and settings, the GPU has relatively little work per frame, so the CPU must generate draw calls and update game state quickly enough to hit that frame rate. The 5.30 GHz boost clock is clearly sufficient, as the FPS continues to scale with settings increases rather than plateauing due to CPU limitations.

The CPU's single-thread benchmark scores — 600 in Cinebench R15 and 5,960 in Cinebench R23 — place it in the 93rd percentile of all processors. These scores, combined with the 42,216 multi-core R23 result, indicate a balanced part that excels in both lightly-threaded and heavily-threaded workloads. For Downfall, a horror game likely built on an engine that uses a few primary threads for gameplay logic and physics, the 5.30 GHz boost is more relevant than the 20-core count.

The nearest CPU rivals in the database — the EPYC 4464P, EPYC 7343, and EPYC 9124 — all score within 0.7% of the Ultra 7 265's average, yet they are server processors with different power characteristics (the Ultra 7 265 has a 65 W TDP). This suggests that for Downfall, the CPU's raw computational throughput matters less than its ability to maintain high clock speeds under gaming loads, which the measured FPS data confirms through consistent performance across all resolutions.

FAQ

*Q: What is the best resolution for this combo in Downfall?*

A: The data shows that 2560x1440 at Medium settings delivers 264 FPS average with a 225 FPS minimum, which offers both high frame rates and reasonable visual quality. At 4K, High settings provide 135 FPS, making it viable for users prioritizing resolution over frame rate.

Q: Does the RTX 5090's 32 GB VRAM make a difference in this game?

A: The measured results show no evidence of VRAM limitations — even at 4K Ultra (84 FPS average), the game runs smoothly, suggesting 32 GB exceeds Downfall's requirements. The GPU's 1.79 TB/s bandwidth ensures memory throughput keeps pace with rendering demands.

Q: How does this combo compare to other configurations in the database?

A: This pairing ranks 56th out of 2,928 tested combos, placing it in the top 2% of all configurations. The GPU's nearest rival (Tesla P100) scores 0.3% lower in synthetic benchmarks, while the CPU is within 0.7% of its closest competitors.

Q: Is the CPU or GPU the limiting factor at 1080p?

A: At 1080p Low, the 452 FPS result approaches CPU limits, but the uniform scaling across settings (from 452 to 190 FPS going Low to Ultra) indicates the GPU remains the primary constraint. The CPU's 5.30 GHz boost clock keeps it ahead of GPU demands.

Q: What frame rate can users expect at 4K Ultra?

A: The measured average is 84 FPS at 3840x2160 with Ultra settings. While playable, this is the only configuration in the dataset that drops below 100 FPS, making High settings (135 FPS) a better choice for smoother motion.

Q: How does the Core Ultra 7 265's performance compare to its rivals?

A: The CPU's average benchmark score of 64,640 is 0.3% lower than the AMD EPYC 4464P (64,823) and 0.7% lower than the EPYC 9124 (65,104), but 0.3% higher than the Core Ultra 7 265F (64,438). These differences are within measurement noise.

Measured FPS Breakdown

At 1920x1080, the full settings range shows: Low averages 452 FPS, Medium averages 370 FPS with a minimum of 314 FPS and maximum of 425 FPS, High averages 303 FPS, and Ultra averages 190 FPS. The Medium preset provides the only recorded minimum/maximum range at this resolution, indicating the other presets may have less variance or the data collection captured fewer frames.

Moving to 2560x1440, the results shift downward: Low averages 323 FPS, Medium averages 264 FPS with a minimum of 225 FPS and maximum of 304 FPS, High averages 217 FPS, and Ultra averages 135 FPS. The Medium preset again shows the tightest frame time distribution, with a 79 FPS spread between minimum and maximum.

At 3840x2160, the highest resolution tested, the averages are: Low at 201 FPS, Medium at 164 FPS with a minimum of 140 FPS and maximum of 189 FPS, High at 135 FPS, and Ultra at 84 FPS. The Medium preset's 49 FPS range at 4K is proportionally similar to the 79 FPS range at 1440p, suggesting consistent frame pacing characteristics across resolutions for this settings level.

The data reveals that Medium is the only preset with complete minimum/maximum measurements across all three resolutions, which may indicate it represents the most stable configuration for this hardware. The gap between Medium and High at each resolution — 67 FPS at 1080p, 47 FPS at 1440p, and 29 FPS at 4K — narrows as resolution increases, reinforcing that GPU workload becomes the dominant factor at higher pixel counts. Ultra settings consistently deliver roughly half the frame rate of Low settings at every resolution, demonstrating the significant cost of maximum visual fidelity in Downfall.