Minimum

Minimum is a 2014 third-person shooter that, when paired with the AMD Ryzen 7 5700G and NVIDIA GeForce RTX 5090 D, produces a dataset dominated by extreme GPU headroom. The measured FPS values across three resolutions and four quality settings reveal a configuration that is almost never constrained by the graphics card, shifting the analytical focus to the CPU’s frame pacing limits and the practical boundaries of the game engine. The data shows a system capable of triple-digit averages even at 4K, but the scaling patterns between settings and resolutions indicate where the bottlenecks actually lie.

Resolution Scaling

The frame rate drop from 1080p to 4K is surprisingly modest, which is the first major indicator that the RTX 5090 D is not the limiting factor. At High settings, the average FPS moves from 276 at 1920x1080 to 178 at 2560x1440, then down to 96 at 3840x2160. That represents a 35.5% reduction from 1080p to 1440p and a further 46.1% reduction from 1440p to 4K. While these are substantial percentage drops, the absolute values remain so high that the GPU is never struggling.

The Low preset tells a clearer story. At 1080p Low, the system hits 350 FPS average. At 1440p Low, it drops to 277 FPS. At 4K Low, it still maintains 149 FPS. The scaling from 1440p to 4K here is a 46.2% reduction, nearly identical to the High preset’s drop. This consistency across presets suggests the GPU is scaling with pixel count as expected, but the starting point is so high that even a 4K Low setting produces nearly 150 FPS.

The Medium preset offers the only complete data with minimum and maximum FPS values. At 1080p Medium, the average is 318 FPS with a minimum of 270 and maximum of 366. At 1440p Medium, the average drops to 212 FPS, with a range of 180 to 244. At 4K Medium, the average falls to 114 FPS, with a range of 97 to 131. The minimum-to-average gap widens slightly at higher resolutions, from 48 FPS at 1080p to 32 FPS at 1440p and 17 FPS at 4K, indicating that frame pacing becomes more consistent as the GPU load increases.

The Ultra preset shows the steepest resolution scaling. At 1080p Ultra, the average is 210 FPS. At 1440p Ultra, it falls to 136 FPS, a 35.2% drop. At 4K Ultra, it collapses to 73 FPS, a 46.3% drop from 1440p. This pattern reveals that Ultra settings impose a significant fixed cost on the GPU, but even at 4K Ultra, the system remains above 60 FPS, which is playable for a third-person shooter.

The limiting component is clearly the CPU at lower resolutions and settings. At 1080p Low, the 350 FPS average suggests the Ryzen 7 5700G is approaching its single-thread ceiling. The 3DMark single-thread score of 899 and Cinebench R23 single-core score of 1494 indicate moderate per-core performance, which aligns with the observed plateau. The GPU is so powerful that it can render frames faster than the CPU can issue draw calls, creating a CPU-bound scenario at 1080p and 1440p Low.

GPU Role

The RTX 5090 D is a Blackwell 2.0 architecture GPU with 32 GB of GDDR7 memory on a 512-bit bus, providing 1.79 TB/s of bandwidth. Its 21,760 shading units and 176 ROPs deliver a pixel rate of 423.6 GPixel/s and a texture rate of 1,636.8 GTexel/s. The base clock is 2017 MHz with a boost clock of 2407 MHz, and memory operates at 1750 MHz with 28 Gbps effective. These specifications place it in the 92nd percentile of all GPUs, with an average benchmark score of 77,712.

The nearest rivals in the GPU benchmark database are the AMD Radeon RX 6650M XT (76,904, deltaPct 1.1), AMD Radeon RX 6850M XT (78,940, deltaPct -1.6), NVIDIA Tesla P100 PCIe 12 GB (79,396, deltaPct -2.1), and NVIDIA Tesla P100 PCIe 16 GB (79,605, deltaPct -2.4). This means the RTX 5090 D is statistically tied with these mobile and datacenter parts in synthetic benchmarks, despite its desktop positioning, but in Minimum, the measured FPS tells a different story.

The VRAM capacity of 32 GB is far beyond what Minimum requires, even at 4K Ultra. The game, released in 2014, likely uses less than 8 GB of VRAM at any setting. This oversupply means memory bandwidth is never a constraint. The 1.79 TB/s bandwidth is more than sufficient to feed the shading units at any resolution. The GPU’s role in this combo is to provide near-unlimited headroom, such that the measured FPS is governed by engine limits and CPU performance rather than GPU throughput.

The FP32 throughput of 104.8 TFLOPS and FP16 of 104.8 TFLOPS (1:1) further underscore the GPU’s capabilities. In a game like Minimum, which relies on standard rasterization rather than ray tracing or tensor core workloads, the RTX 5090 D’s 170 RT cores and 680 tensor cores are largely idle. The passmark DirectX 11 score of 371 and DirectX 12 score of 219 indicate strong API performance, but the game’s older engine likely benefits more from raw shader throughput than advanced features.

FAQ

Q: Why does the frame rate stay above 60 FPS even at 4K Ultra?

A: The RTX 5090 D’s 32 GB GDDR7 memory and 1.79 TB/s bandwidth, combined with 104.8 TFLOPS FP32 performance, provide massive headroom for a 2014 game. The measured 73 FPS average at 3840x2160 Ultra is the lowest result in the dataset, but it remains playable, indicating the GPU is never the bottleneck.

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

A: The 350 FPS average at 1920x1080 Low is likely CPU-bound. The Ryzen 7 5700G’s 3DMark single-thread score of 899 and Cinebench R23 single-core score of 1494 suggest a per-core ceiling that caps frame rates around this level, while the GPU could render faster if the CPU could issue more frames.

Q: What is the best preset for consistent frame pacing at 1440p?

A: The Medium preset at 2560x1440 shows a 212 FPS average with a minimum of 180 and maximum of 244. The 32 FPS gap between minimum and average is smaller than the 48 FPS gap at 1080p Medium, indicating better frame pacing at higher resolution.

Q: How does the combo rank against other tested systems?

A: The AMD Ryzen 7 5700G + NVIDIA GeForce RTX 5090 D combo ranks 53rd out of 3,712 tested combinations in Minimum, placing it in the top 1.4% of all systems tested for this game.

Q: Does the GPU’s 32 GB VRAM matter for this game?

A: No. Minimum, released in 2014, does not require anywhere near 32 GB of memory. The oversupply means VRAM capacity has zero impact on the measured results; the bandwidth and shader throughput are the relevant factors.

Q: Why does the Ultra preset show such a large drop from 1440p to 4K?

A: At 4K Ultra, the average FPS is 73, compared to 136 at 1440p Ultra, a 46.3% reduction. This is the largest relative drop in the dataset, indicating that Ultra settings impose a fixed rendering cost that scales heavily with pixel count, unlike Low or Medium settings.

How This Combo Ranks

The combo rank of 53 out of 3,712 tested combinations places this system in the 98.6th percentile for Minimum. This is an exceptional position, but the context matters. The Ryzen 7 5700G is a 2021 CPU with a 79th percentile ranking among all CPUs, while the RTX 5090 D is a 2025 GPU with a 92nd percentile ranking. The combination of a mid-tier CPU with a top-tier GPU creates a system that excels in GPU-bound scenarios but is held back in CPU-bound ones.

The CPU’s nearest rivals in synthetic benchmarks are the Intel Core i9-9900K (avgScore 27,097, deltaPct -0.2), AMD Ryzen AI 7 445 (avgScore 26,936, deltaPct 0.4), Intel Core i5-14400T (avgScore 27,166, deltaPct -0.4), and Intel Core i7-1370P (avgScore 26,900, deltaPct 0.6). These deltas are all within 0.6%, meaning the 5700G is statistically identical to these CPUs in overall benchmark performance. However, in Minimum, the 5700G’s 8 cores and 16 threads with a boost clock of 4.60 GHz provide enough single-thread performance to sustain high frame rates, though not enough to fully unlock the GPU.

The GPU’s nearest rivals are all within 2.4% in synthetic scores, but the RTX 5090 D’s 92nd percentile ranking and 77,712 average benchmark score place it far above the median GPU. In Minimum, the GPU’s performance is so high that it never appears in the measured FPS as a limiting factor. The rank of 53rd reflects the CPU’s ability to keep up with the GPU at lower resolutions, but at 4K Ultra, the system still produces 73 FPS, which is higher than most systems with weaker GPUs or stronger CPUs.

Measured FPS Breakdown

At 1920x1080, the system achieves its highest frame rates. Low settings produce a 350 FPS average. Medium settings yield a 318 FPS average with a minimum of 270 and maximum of 366. High settings drop to 276 FPS average. Ultra settings fall to 210 FPS average. The spread from Low to Ultra is 140 FPS, indicating that quality settings have a significant impact at this resolution.

At 2560x1440, the frame rates remain high but show more differentiation. Low settings produce 277 FPS average. Medium settings yield 212 FPS average with a minimum of 180 and maximum of 244. High settings drop to 178 FPS average. Ultra settings fall to 136 FPS average. The spread from Low to Ultra is 141 FPS, nearly identical to 1080p, suggesting the GPU scales linearly with pixel count while maintaining the same settings-based delta.

At 3840x2160, the frame rates finally show signs of GPU load. Low settings produce 149 FPS average. Medium settings yield 114 FPS average with a minimum of 97 and maximum of 131. High settings drop to 96 FPS average. Ultra settings fall to 73 FPS average. The spread from Low to Ultra is 76 FPS, which is roughly half the spread seen at lower resolutions. This compression indicates that the GPU is becoming more constrained at 4K, but the absolute values remain high.

The Medium preset is the only one with complete minimum and maximum data across all three resolutions. At 1080p, the minimum of 270 is 84.9% of the average. At 1440p, the minimum of 180 is 84.9% of the average. At 4K, the minimum of 97 is 85.1% of the average. This consistency suggests that frame pacing is stable across resolutions, with the minimum FPS always hovering around 85% of the average.

Settings Recommendations

The data indicates that the Medium preset offers the best balance of visual quality and frame rate consistency. At 1080p Medium, the system produces 318 FPS average, which is only 9.1% lower than Low but provides better visuals. At 1440p Medium, the 212 FPS average is well above any display refresh rate currently available, and the minimum of 180 FPS ensures no dips below 144 Hz. At 4K Medium, the 114 FPS average with a minimum of 97 FPS keeps the experience smooth on high-refresh 4K displays.

The High preset is a reasonable choice for those prioritizing visual fidelity without sacrificing playability. At 1440p High, the 178 FPS average is still above 144 Hz. At 4K High, the 96 FPS average is above 90 Hz, which is acceptable for a third-person shooter. However, the Ultra preset at 4K drops to 73 FPS, which may cause noticeable frame pacing issues on 120 Hz displays, though it remains above 60 FPS.

For competitive play, the Low preset at 1080p or 1440p is recommended. The 350 FPS at 1080p and 277 FPS at 1440p provide the lowest input latency, and the visual downgrade is acceptable in a fast-paced shooter. The key takeaway is that the Medium preset provides the best experience per the measured rows, as it maintains high frame rates with a minimum FPS that never drops below 97 at any resolution.

CPU Role

The AMD Ryzen 7 5700G is a Zen 3 (Cezanne) architecture processor on a 7 nm process, with 8 cores and 16 threads. Its base clock is 3.80 GHz with a boost clock of 4.60 GHz, and it has a 65 W TDP. The cache hierarchy includes 64 KB of L1 per core, 512 KB of L2 per core, and 16 MB of L3 cache. Memory support is DDR4 dual-channel with 51.2 GB/s bandwidth. The CPU’s integrated Radeon Vega 8 graphics are irrelevant in this configuration, as the RTX 5090 D handles all rendering.

In synthetic benchmarks, the 5700G scores 12,830.5 in Cinebench R23 multicore and 1,494 in single-core. Its 3DMark scores range from 899 single-thread to 6,629 for 16 threads. The Passmark multithread score is 24,419, and single-thread is 3,283. These numbers place the CPU in the 79th percentile, with an average benchmark score of 27,051, statistically tied with the Intel Core i9-9900K.

In Minimum, the CPU’s role is to feed the GPU with draw calls. At 1080p Low, the 350 FPS average is likely near the CPU’s maximum frame generation capability. The single-thread performance, measured at 899 in 3DMark and 1,494 in Cinebench R23, determines this ceiling. The 8-core, 16-thread configuration provides ample multi-threaded headroom, but the game’s engine likely relies on a few primary threads, making single-core performance the critical factor.

The boost clock of 4.60 GHz is respectable but not exceptional for a 2021 processor. The 16 MB L3 cache helps with game data locality, and the 51.2 GB/s memory bandwidth is sufficient for the CPU’s needs. The data shows that the CPU becomes the limiting factor only at 1080p Low, where the GPU could render faster if the CPU were stronger. At 1440p and above, the GPU load increases enough that the CPU has room to keep up, as evidenced by the consistent scaling patterns.