Minimum

Minimum is a 2014 third-person shooter that, on paper, should be trivial for modern hardware, but the measured data for the Intel Core Ultra X7 368H and NVIDIA GeForce RTX 5090 combo tells a more nuanced story about where bottlenecks appear. This pairing of a 16-core Panther Lake mobile CPU and a flagship desktop-class GPU produces frame rates ranging from playable at 4K Ultra to exceptionally high at 1080p Low, but the scaling pattern reveals that the game's engine and the CPU's specific characteristics create a ceiling that even the RTX 5090 cannot push past at lower resolutions.

Resolution Scaling

The frame rate progression from 1080p to 4K follows a predictable curve, but the magnitude of the drops reveals which component is governing performance. At High settings, the average FPS falls from 188 at 1920x1080 to 121 at 2560x1440—a drop of 67 FPS—and then further down to 65 at 3840x2160. This represents a 65% reduction from 1080p to 4K High, which is substantial but not purely GPU-linear, suggesting that something other than raw pixel throughput is at play. If the RTX 5090 were the sole limiter, the 4K score would typically be closer to a quarter of the 1080p output, but here it is roughly a third.

Looking at Low settings, the pattern is even more revealing. The jump from 293 FPS at 1080p to 189 FPS at 1440p is a 104 FPS loss, and then to 102 FPS at 4K is another 87 FPS loss. The 1080p Low result of 293 FPS appears to hit a hard ceiling—likely a CPU or engine limitation—because the GPU is capable of far higher frame rates at such low graphical load. The fact that 1080p Low only reaches 293 FPS, while 1080p High reaches 188 FPS, indicates that the CPU is already the primary constraint at 1080p, with the GPU only becoming the limiter as resolution increases and pixel work grows.

The Medium settings provide the only complete data set with min and max FPS values. At 1080p Medium, the average is 224 FPS with a minimum of 190 and maximum of 258. At 1440p Medium, the average drops to 144 FPS with a range of 123 to 166. At 4K Medium, the average falls to 78 FPS with a minimum of 66 and maximum of 89. The consistency of the min-to-max range across resolutions—roughly 68-70 FPS spread—suggests the frame pacing is stable, but the average drops indicate that each resolution step adds enough pixel load to shift the bottleneck from CPU to GPU. At 4K Medium, the 78 FPS average is close to what one might expect from a pure GPU-bound scenario, whereas at 1080p Medium, the 224 FPS average is likely CPU-capped.

Ultra settings show the steepest drop-off. From 143 FPS at 1080p to 92 FPS at 1440p and then to 50 FPS at 4K, the scaling is more aggressive than High settings. This suggests that Ultra settings introduce effects that disproportionately tax the GPU, making the 4K Ultra result of 50 FPS the only truly demanding scenario in this game for the RTX 5090. The data indicates that for Minimum, the CPU is the limiting factor at 1080p across all settings, while the GPU takes over as the primary constraint at 1440p and 4K, with the crossover point occurring somewhere between 1080p and 1440p depending on settings.

GPU Role

The RTX 5090 in this combo is a 32 GB GDDR7 card with a 512-bit memory bus delivering 1.79 TB/s of bandwidth, and its clock speeds are listed at 2017 MHz base and 2407 MHz boost. With 21,760 shading units and 176 ROPs, this is a GPU that should obliterate a 2014 title, yet the measured FPS at 4K Ultra is only 50. This result is not a reflection of the GPU's raw capability but rather the game's engine limits and the CPU's ability to feed the GPU with draw calls. The GPU's passmark G3D score of 39,650 and its 92nd percentile ranking among all GPUs confirm it is a top-tier part, but Minimum's older engine does not fully utilize modern parallel architectures.

The memory configuration is notable for this game. While 32 GB of VRAM is overkill for Minimum—a title from 2014 that likely uses less than 4 GB even at 4K—the bandwidth of 1.79 TB/s ensures that texture streaming and memory access are never bottlenecks. The data supports this: at 4K Medium, the average FPS of 78 with a minimum of 66 shows that even when the GPU is heavily loaded, it maintains consistent frame delivery. The GPU's 104.8 TFLOPS of FP32 compute is more than sufficient for this game's shader complexity, so the 4K Ultra 50 FPS result must stem from engine-level inefficiencies rather than GPU resource exhaustion.

The RTX 5090's transistor count of 92,200 million on a 5 nm process with a die size of 750 mm² gives it a 122.9M transistors per mm² density. This physical capacity translates to benchmark scores like 3DMark Steel Nomad DX12 at 18,355 and Geekbench Vulkan at 376,728, which are class-leading results. However, in Minimum, the GPU's performance is constrained by how quickly the CPU can process game logic and issue rendering commands. At 1080p Low, the 293 FPS average is likely close to the engine's maximum frame rate, and the GPU is sitting idle waiting for work—a classic CPU-bound scenario where the RTX 5090's immense power is underutilized.

The GPU's nearest rivals in the benchmark database—the NVIDIA Tesla P100 PCIe 16 GB with a 0.3% higher average score, and the AMD Radeon RX 6850M XT at 1.1% higher—are not relevant to gaming performance in Minimum, but they do contextualize the RTX 5090's position. The data shows that the GPU is not the limiting factor at 1080p, and only at 4K Ultra does it start to approach its performance ceiling in this specific title. For a game like Minimum, the RTX 5090 is effectively over-provisioned, but it ensures that resolution scaling is smooth and that even the heaviest settings remain playable.

CPU Role

The Intel Core Ultra X7 368H is a 16-core, 16-thread processor with no hyperthreading, based on the Panther Lake architecture manufactured on Intel's 3 nm process. Its base clock of 2.00 GHz and boost clock of 5.00 GHz are the key metrics for game performance, and the benchmark data shows this CPU is positioned in the 87th percentile of all CPUs with an average benchmark score of 40,518. In Minimum, the CPU's role is critical because the game's engine appears to be single-thread or lightly-threaded, meaning that the 16 cores are less important than the boost clock behavior.

The Cinebench R23 multi-core score of 28,221 and single-core score of 3,984 provide context. The single-core result is strong, and the 5.00 GHz boost clock should handle Minimum's logic-heavy scenes well. However, the 1080p Low result of 293 FPS suggests the CPU is still the ceiling. The nearest rivals—the Intel Core 7 253PE at -0.1% delta, the Intel Core 5 223PE at -0.2%, and the AMD Ryzen 9 7940H at +0.2%—all have nearly identical average scores, indicating that this CPU is not uniquely positioned for gaming workloads. The Xeon 6507P at +0.2% delta is a server part, further suggesting that raw CPU compute is not the differentiator in this title.

The CPU's memory support is LPDDR5X with dual-channel configuration and 153.6 GB/s bandwidth. This is relevant because Minimum's engine may be sensitive to memory latency and bandwidth, especially at high frame rates where the CPU must process many draw calls per second. The 18 MB of shared L3 cache and 2.5 MB per-core L2 cache are adequate for the game's data footprint, but the lack of hyperthreading means that if the engine spawns more than 16 threads, there could be contention. The Passmark multi-thread score of 32,956 and single-thread score of 4,005 show a balanced profile, but the physics score of 2,857 is notably lower, which could impact a third-person shooter with physics-based interactions.

The CPU's integrated graphics (Arc B390) are irrelevant for this combo since the RTX 5090 handles rendering, but the CPU's PCIe Gen 5 with only 4 lanes from the CPU is a potential constraint for GPU communication. This could explain why the frame rates at 1080p do not scale as high as expected—the 4-lane connection may limit data transfer between the CPU and GPU, even if the RTX 5090 uses a PCIe 5.0 x16 interface. The data does not provide a definitive answer, but the 293 FPS ceiling at 1080p Low is suspiciously low for a 5.00 GHz CPU paired with the fastest consumer GPU, suggesting that either the game engine's threading model or the CPU's I/O configuration is the bottleneck.

Measured FPS Breakdown

At 1920x1080, the measured averages are 293 FPS at Low, 224 FPS at Medium, 188 FPS at High, and 143 FPS at Ultra. The Medium setting provides the only min/max data at this resolution: 190 min and 258 max, which is a 68 FPS spread indicating consistent frame delivery. The gap between Low and Medium is 69 FPS, while the gap between Medium and High is 36 FPS, and between High and Ultra is 45 FPS. This nonlinear scaling suggests that Low settings remove enough visual effects to let the CPU push near its limit, while each higher setting adds GPU work that progressively shifts the bottleneck.

At 2560x1440, the averages are 189 FPS at Low, 144 FPS at Medium, 121 FPS at High, and 92 FPS at Ultra. The Medium min/max is 123 to 166, a 43 FPS spread that is tighter than at 1080p, indicating that the GPU is starting to take over and smooth out frame pacing. The drop from 1080p to 1440p is 104 FPS at Low, 80 FPS at Medium, 67 FPS at High, and 51 FPS at Ultra. The proportional drops are larger at lower settings, confirming that the CPU bottleneck at 1080p is being relieved as resolution increases.

At 3840x2160, the averages are 102 FPS at Low, 78 FPS at Medium, 65 FPS at High, and 50 FPS at Ultra. The Medium min/max is 66 to 89, a 23 FPS spread that is the tightest of all resolutions, showing that at 4K the GPU is fully engaged and frame delivery is stable. The drop from 1440p to 4K is 87 FPS at Low, 66 FPS at Medium, 56 FPS at High, and 42 FPS at Ultra. Interestingly, the absolute drops at 4K are larger than the 1440p drops for Low and Medium, but smaller for High and Ultra, suggesting that Ultra settings at 4K are hitting a GPU compute limit that reduces the resolution sensitivity.

The complete dataset shows that the combo achieves playable frame rates at all settings and resolutions, with the worst case being 4K Ultra at 50 FPS average. For a third-person shooter, 50 FPS is borderline but playable, while 65 FPS at 4K High is comfortable. The 1080p Low result of 293 FPS is the standout, but it is not usable for competitive play due to the visual sacrifices. The sweet spot appears to be 1440p High at 121 FPS or 4K Medium at 78 FPS, both of which balance visual quality with high refresh rate capability.

How This Combo Ranks

The combo rank for this pairing in Minimum is 581 out of 2,939 tested combinations, placing it in the top 19.8% of all tested systems. This ranking is lower than one might expect given the RTX 5090's 92nd percentile GPU ranking and the CPU's 87th percentile, indicating that the combination underperforms relative to its component strengths. The rank suggests that other CPU-GPU pairings achieve higher frame rates in this game, likely because Minimum does not scale with the RTX 5090's massive compute resources and instead favors CPUs with higher single-thread performance or more favorable memory configurations.

The rank of 581 out of 2,939 means that 2,358 other combos perform better in Minimum, which is a significant number. This is not a reflection of poor hardware but rather a mismatch between the game's engine requirements and the combo's characteristics. The RTX 5090's 32 GB VRAM and 512-bit bus are wasted on a game that likely uses a fraction of these resources, and the CPU's 16 cores are underutilized if the game is single-threaded. The combo's performance would likely be much higher in a modern, multi-threaded title that can exploit the GPU's 21,760 shading units and the CPU's multi-core strength.

The data indicates that this combo is overkill for Minimum, and the rank reflects that the game cannot fully utilize the hardware. The nearest rival CPU scores are all within 0.2% of each other, suggesting that the CPU choice has minimal impact on the game's performance, while the GPU's nearest rivals are also within 1.4%, indicating that GPU choice matters more but still not decisively. The 581st rank is a warning to builders that flagship hardware does not automatically translate to top-tier performance in older titles, and that game-specific optimization is often more important than raw component power.

FAQ

Q: What is the highest average FPS this combo achieves in Minimum?

A: The highest average FPS is 293 at 1920x1080 with Low settings. At Medium settings, the average is 224 FPS with a minimum of 190 and maximum of 258. The 293 FPS result appears to be a CPU or engine ceiling, as the RTX 5090 is likely capable of much higher frame rates at such low graphical load.

Q: Is 4K Ultra playable with this combo?

A: Yes, but just barely. The 4K Ultra average is 50 FPS, which is playable for a single-player third-person shooter but not ideal for competitive play. For a smoother experience, 4K High delivers 65 FPS average, and 4K Medium delivers 78 FPS with a minimum of 66 FPS, ensuring no noticeable stutters.

Q: Which component is the bottleneck at 1080p resolution?

A: The CPU is the bottleneck at 1080p. Evidence includes the 293 FPS ceiling at Low settings, which is far below what the RTX 5090 could theoretically produce, and the fact that dropping from High to Low only increases FPS by 105 (from 188 to 293), indicating the CPU cannot feed the GPU fast enough. The CPU's 16 cores without hyperthreading and 4-lane PCIe Gen 5 connection may contribute to this limitation.

Q: How does this combo rank compared to other tested systems in Minimum?

A: This combo ranks 581st out of 2,939 tested combinations, placing it in the top 19.8%. This is lower than expected given the GPU's 92nd percentile and CPU's 87th percentile rankings, indicating that Minimum does not scale well with this hardware. The game likely favors different CPU characteristics or memory configurations.

Q: What is the best settings/resolution balance for this combo?

A: The data suggests 1440p High at 121 FPS average is the best balance of visual quality and speed. For higher fidelity, 4K Medium at 78 FPS average with a 66 FPS minimum is also excellent. 1080p Ultra at 143 FPS is playable, but 1440p High provides more detail with only a 22 FPS penalty.

Q: Does the RTX 5090's 32 GB VRAM matter for this game?

A: No, the 32 GB GDDR7 memory is far more than Minimum requires. The game is from 2014 and likely uses a fraction of this capacity even at 4K Ultra. The 1.79 TB/s bandwidth ensures texture streaming is never an issue, but the VRAM size has no practical impact on the measured FPS in this title.