Evolve

# Evolve with AMD Ryzen 5 4600G + NVIDIA GeForce RTX 5090 D

This pairing combines a 6-core/12-thread AMD Ryzen 5 4600G (Zen 2, Renoir, 7nm) with NVIDIA’s flagship RTX 5090 D (Blackwell 2.0, 32GB GDDR7) in a 2015 FPS title. The benchmark data reveals a system that is heavily GPU-limited at lower resolutions but becomes increasingly balanced as resolution climbs, though the CPU’s modest single-thread performance leaves headroom even at 4K.

Resolution Scaling

The measured FPS data shows a clear pattern: performance drops progressively as resolution increases, but the magnitude of that drop varies dramatically by settings preset. At High settings, the average frame rate falls from 193 FPS at 1920x1080 to 165 FPS at 2560x1440 (a 14.5% reduction), then to 94 FPS at 3840x2160 (a 43% reduction from 1440p, or 51% from 1080p). This steep decline at 4K indicates the RTX 5090 D’s rendering workload is scaling heavily with pixel count, suggesting the GPU becomes the primary bottleneck at 3840x2160 even with this massive 32GB frame buffer.

At Low settings, the pattern is similar but less extreme: 232 FPS at 1080p, 203 FPS at 1440p (12.5% drop), and 154 FPS at 4K (24.1% drop from 1440p). The smaller percentage decline at Low settings suggests that even at 4K, the CPU is still contributing meaningfully to the frame rate ceiling. The gap between Low and High at 4K (154 vs 94 FPS) is 60 FPS, whereas at 1080p that same gap is only 39 FPS (232 vs 193). This widening gap at higher resolutions confirms that the RTX 5090 D’s raw power is being more fully utilized as resolution increases, while the Ryzen 5 4600G’s limits become less relevant.

Ultra settings tell a different story. At 1080p Ultra, the combo hits 169 FPS, which is only 24 FPS below High. At 1440p Ultra, it drops to 114 FPS (32.5% below 1080p Ultra), and at 4K Ultra it falls to 65 FPS — a 43% drop from 1440p Ultra and a 61.5% drop from 1080p Ultra. The 4K Ultra result of 65 FPS is the only sub-60-adjacent figure in the entire dataset, and it indicates that at maximum settings with 4K resolution, the combined load of geometry, shading, and pixel processing finally overwhelms what this CPU-GPU pairing can sustain.

The Medium preset provides the only complete min/max FPS data in the pack. At 1080p Medium, the average is 211 FPS with a minimum of 179 and maximum of 243. At 1440p Medium, the average drops to 185 FPS (min 157, max 212), and at 4K Medium it reaches 120 FPS (min 102, max 138). The spread between min and max FPS at each resolution (64 FPS at 1080p, 55 FPS at 1440p, 36 FPS at 4K) narrows as resolution increases, which suggests that frame time variance is more influenced by CPU-side fluctuations at lower resolutions where the GPU is less stressed.

FAQ

Q: Is this CPU-GPU pairing bottlenecked by the processor at 1080p?

A: Yes, the data strongly indicates CPU limitation at 1080p. At Low settings, 1080p (232 FPS) is only 14.3% higher than 1440p (203 FPS), which is a much smaller gain than the 32.5% pixel count increase. Even at High settings, 1080p (193 FPS) is only 17% faster than 1440p (165 FPS), despite rendering 44% fewer pixels. This compressed scaling suggests the Ryzen 5 4600G’s single-thread performance (3DMark single-thread score of 726, Cinebench R23 single-core of 1919) is capping frame rates.

Q: Can this combo maintain 144 FPS at 4K?

A: No, not in any tested preset. The highest 4K result is 154 FPS at Low settings, but High drops to 94 FPS, Medium to 120 FPS, and Ultra to 65 FPS. Even the best-case Low preset at 4K cannot sustain 144 FPS consistently, and the Medium preset’s minimum of 102 FPS falls well below the 144 FPS threshold.

Q: How does the RTX 5090 D’s 32GB VRAM impact Evolve?

A: The 32GB GDDR7 frame buffer (with 1.79 TB/s bandwidth) is far beyond what this 2015 title requires. The benchmark data shows no evidence of VRAM-related stuttering — the minimum FPS figures at Medium settings are proportional to averages (e.g., 179 min vs 211 avg at 1080p), indicating the GPU never runs out of memory. The 512-bit bus and 1.79 TB/s bandwidth allow the GPU to feed its 21760 shading units without constraint.

Q: Is the Ryzen 5 4600G’s 8MB L3 cache a limiting factor?

A: The small shared L3 cache (8MB) compared to modern desktop chips could contribute to the CPU bottleneck at 1080p, but the benchmark data alone cannot isolate this. The CPU’s 3DMark scores (4863 for 16 threads, 1430 for 2 threads) and PassMark single-thread score of 2653 indicate moderate performance that aligns with the observed 1080p plateau around 193-232 FPS depending on settings.

Q: What is the best settings preset for this combo at 1440p?

A: High settings at 1440p delivers 165 FPS average, which is a strong balance. Ultra drops to 114 FPS — still playable but a 31% reduction. Medium at 1440p (185 FPS) offers only 12% more than High, making High the more visually efficient choice. For competitive play, Low at 1440p hits 203 FPS, but High provides the best quality-to-FPS tradeoff.

Q: How does this combo rank among all tested systems for Evolve?

A: It ranks 208th out of 2872 tested combos, placing it in the top 7.3% of all configurations. This is respectable, but given that the RTX 5090 D sits in the 92nd percentile of all GPUs, the ranking suggests the CPU is holding the combo back from the top tier — a system with a stronger processor would likely rank higher.

How This Combo Ranks

The combo ranks 208 out of 2872 tested combinations for Evolve, which places it in the 92.8th percentile of all tested systems. This is a strong result, but it’s worth contextualizing against the individual components’ positions. The RTX 5090 D has a percentile of 92 against all GPUs, while the Ryzen 5 4600G has a percentile of 71 against all CPUs. The combo’s rank of 208 is closer to the GPU’s percentile than the CPU’s, suggesting that in Evolve specifically, the graphics card is doing most of the heavy lifting for the overall ranking.

Looking at the CPU’s nearest rivals, the Ryzen 5 4600G (avg benchmark score 17507) sits within 0.8% of the AMD Ryzen 3 8300GE (17614), 0.6% of the Intel Core 7 150U (17395), and 0.1% of the AMD Ryzen 3 PRO 5355GE (17482). These are all close scores, but they’re all lower-tier chips compared to what would typically pair with an RTX 5090 D. The GPU’s nearest rivals include the AMD Radeon RX 6650M XT (76904, 1.1% lower), the RX 6850M XT (78940, 1.6% higher), and two Tesla P100 variants — none of which are in the same performance class as the RTX 5090 D’s 77712 average score.

The 208th rank out of 2872 combos means there are 207 combinations that beat this setup in Evolve, and 2664 that it beats. Given the GPU’s top-tier status, the fact that it doesn’t crack the top 100 suggests that CPU-bound scenarios in Evolve (likely at 1080p and 1440p) are dragging the average down. A system with a higher single-thread CPU would likely see this combo rank significantly higher.

Measured FPS Breakdown

At 1920x1080, the combo delivers its highest frame rates across all settings: 232 FPS at Low, 211 FPS at Medium (min 179, max 243), 193 FPS at High, and 169 FPS at Ultra. The step from Ultra to High is 24 FPS, from High to Medium is 18 FPS, and from Medium to Low is 21 FPS. This relatively even scaling (roughly 20 FPS per settings tier) indicates that at 1080p, the GPU has enough headroom that settings changes translate directly into FPS changes, but the overall ceiling is set by the CPU — note that even Low settings can’t exceed 232 FPS.

At 2560x1440, frame rates remain high: 203 FPS at Low, 185 FPS at Medium (min 157, max 212), 165 FPS at High, and 114 FPS at Ultra. The gap between High and Ultra widens to 51 FPS, while the gap between Low and Medium narrows to 18 FPS. This suggests that Ultra settings at 1440p start to push the GPU harder, while the lower presets are still partially CPU-limited. The 1440p Medium minimum of 157 FPS is still far above typical 144Hz display refresh rates, making this resolution well-suited for high-refresh gaming.

At 3840x2160, the picture changes dramatically. Low settings produce 154 FPS, Medium produces 120 FPS (min 102, max 138), High produces 94 FPS, and Ultra produces just 65 FPS. The scaling from Low to Ultra is now severe: a 58% reduction in FPS from Low to Ultra. The Medium preset’s minimum of 102 FPS means even that setting can’t guarantee a 100+ FPS experience consistently. At 4K Ultra, the 65 FPS average is the only result below 100 FPS in the entire dataset, and it’s the clearest sign that this resolution/settings combination finally stresses the GPU to its limits — or rather, the limits of what this CPU can feed it.

GPU Role

The RTX 5090 D is a Blackwell 2.0 architecture GPU built on a 5nm process at TSMC, with 92,200 million transistors on a 750 mm² die. Its 32GB of GDDR7 memory operates across a 512-bit bus, delivering 1.79 TB/s of bandwidth. The GPU runs at a base clock of 2017 MHz and boosts to 2407 MHz, with memory at 1750 MHz (28 Gbps effective). Its 21760 shading units, 680 TMUs, and 176 ROPs give it a pixel rate of 423.6 GPixel/s and a texture rate of 1,636.8 GTexel/s, with FP32 performance of 104.8 TFLOPS.

In Evolve, the GPU’s role becomes more pronounced as resolution increases. At 1080p, the FPS deltas between settings (e.g., 232 to 193 FPS from Low to High, a 17% drop) are modest, indicating the GPU is not fully utilized. At 4K, the same Low-to-High transition yields a 34% FPS drop (154 to 94 FPS), showing the GPU is now the dominant factor. The 4K Ultra result of 65 FPS represents the hardest workload in the dataset, and even then, the GPU’s massive compute resources (104.8 TFLOPS FP32) are likely not maxed out — the 65 FPS figure probably reflects a combination of GPU load and CPU-driven draw call limits.

The RTX 5090 D’s 32GB VRAM is irrelevant for Evolve’s actual memory footprint, but the bandwidth (1.79 TB/s) ensures that texture streaming and geometry processing never stall. The GPU’s 92nd percentile ranking (avg benchmark score 77712) places it well above its nearest rivals (RX 6650M XT at 76904, RX 6850M XT at 78940), confirming its top-tier status. Yet in this combo, its performance is capped by the CPU at lower resolutions, with the 1080p High result of 193 FPS representing a fraction of what the GPU could theoretically deliver.

CPU Role

The AMD Ryzen 5 4600G is a 6-core/12-thread processor based on Zen 2 (Renoir) architecture, manufactured on a 7nm process at TSMC with 9,800 million transistors on a 156 mm² die. It runs at a base clock of 3.70 GHz and boosts to 4.20 GHz, with 64KB L1 and 512KB L2 per core, plus 8MB of shared L3 cache. Its memory bandwidth is 51.2 GB/s (dual-channel DDR4), and it has a 65W TDP.

In this game, the CPU’s 3DMark scores tell the story: single-thread 726, 2-thread 1430, 4-thread 2732, 8-thread 4138, and 16-thread 4863. The relatively flat scaling from 8 to 16 threads (4138 to 4863, only 17.5% gain) suggests that Evolve doesn’t scale well beyond 8 threads, and the single-thread score of 726 is the key limiter. Cinebench results reinforce this: R23 multicore of 13593 but single-core of only 1919, with Geekbench single-core at 1460 and PassMark single-thread at 2653.

The CPU’s percentile of 71 against all CPUs is respectable but not elite. Its nearest rivals — the Ryzen 3 PRO 5355GE (17482, +0.1%), Ryzen 3 8300GE (17614, -0.6%), Intel Core 7 150U (17395, +0.6%), and Core i3-14100T (17648, -0.8%) — are all lower-core-count or lower-power parts, indicating the 4600G sits in a mid-range bucket. For Evolve, the CPU’s role is to feed the GPU with draw calls and game logic. At 1080p Low (232 FPS), the CPU is clearly the bottleneck — the GPU could render far more frames, but the CPU can’t issue them fast enough. Even at 4K Low (154 FPS), the CPU may still be limiting, as the 4K High result (94 FPS) shows a 39% drop from 4K Low, which is a larger percentage decline than the pixel count increase would suggest if the GPU were purely limited.

The CPU’s boost clock of 4.20 GHz is modest by modern standards, and its 8MB L3 cache is small for a 6-core part. These factors likely contribute to the observed FPS ceilings around 232 FPS at 1080p Low. For a 2015 game like Evolve, the single-thread performance is the most critical metric, and the 4600G’s 3DMark single-thread score of 726 and PassMark single-thread of 2653 place it in a range that caps frame rates below what the RTX 5090 D could otherwise deliver.