Valorant

Resolution Scaling

The measured FPS data for Valorant shows a surprisingly flat scaling curve across resolutions, which immediately signals that this combo is overwhelmingly CPU-limited at every setting tier. From 1080p Low to 4K Low, the average FPS drops from 540.6 to 474.4, a decline of only 66.2 FPS or roughly 12.2%. That is an exceptionally narrow spread for a game, and it means the RTX 4090 is barely breaking a sweat even at 4K.

The pattern holds across all settings. At High settings, the drop from 1080p (449.5 FPS) to 4K (396.8 FPS) is only 52.7 FPS. At Medium, the spread is 456.0 down to 399.9, a 56.1 FPS difference. Even at Ultra, where you might expect the GPU to finally feel the load, the 4K result (392.0 FPS) is just 48.8 FPS behind the 1080p result (440.8 FPS). The resolution scaling is so mild that it suggests the rendering pipeline is bottlenecked by the CPU's frame submission rate rather than the GPU's pixel throughput.

What this says about the limiting component is clear: the Ryzen 7 5800XT is the ceiling here. The RTX 4090, with its 24 GB of GDDR6X memory and 1.01 TB/s bandwidth, is capable of far higher frame rates than what the CPU can feed it. The fact that dropping resolution from 4K to 1080p only yields a ~12-13% improvement in FPS is textbook CPU-bound behavior. If the GPU were the limiter, the 4K-to-1080p delta would typically be much larger, often 40-60% or more in demanding titles.

Interestingly, the Low preset at 4K (474.4 FPS) outperforms the High preset at 1080p (449.5 FPS). This indicates that even at the highest resolution tested, reducing graphical load frees up enough GPU headroom to push past the CPU limit slightly, but the gains are modest. The data strongly suggests that for competitive play, the extra resolution headroom is effectively free—there is no meaningful penalty for playing at 1440p or 4K versus 1080p, as the CPU is the binding constraint in all cases.

How This Combo Ranks

The AMD Ryzen 7 5800XT + NVIDIA GeForce RTX 4090 combination ranks 402nd out of 1,650 tested combos in Valorant, placing it in the top 24.4% of all configurations tested. This is a strong result, but it is not the absolute top tier, which is telling given the hardware involved. The RTX 4090 is the most powerful consumer GPU in its generation, and the 5800XT is a capable 8-core Zen 3 part, yet the combo does not crack the top 400.

The rank reflects the CPU's single-thread performance ceiling. Valorant is a game that heavily favors high IPC and fast single-core throughput. The 5800XT has a boost clock of 4.80 GHz and scores 959 in the 3DMark single-thread test, which is respectable but not class-leading. The CPU's 86th percentile ranking among all CPUs suggests there are many processors with higher single-thread performance, and those would likely push the combo higher in the rankings.

It is also worth noting the CPU's nearest rivals in the benchmark database: the AMD Ryzen 7 PRO 5755GE (0.1% higher average score), the AMD Ryzen 7 6800HS (0.3% lower), the AMD Ryzen AI 7 PRO 360 (0.3% higher), and the AMD Ryzen 7 7840H (0.3% lower). These are all within a razor-thin margin of each other, indicating that the 5800XT is in a cluster of similarly-performing CPUs. The rank of 402 is likely more a function of the GPU and CPU pairing being measured against combos with faster newer CPUs, not because the 5800XT is weak in absolute terms.

GPU Role

The RTX 4090's role in this pairing is almost entirely about headroom rather than measured load. With 24 GB of GDDR6X memory on a 384-bit bus, the GPU has memory bandwidth (1.01 TB/s) that is vastly underutilized in Valorant. The card's compute capabilities are equally oversized: 82.58 TFLOPS of FP32 performance, 16,384 shading units, and 512 texture mapping units. None of these resources are being taxed by a game like Valorant at these settings.

The clock behavior is worth examining. The GPU has a base clock of 2235 MHz and a boost clock of 2520 MHz. In a CPU-bound scenario, the GPU will often downclock or run at lower utilization because it is waiting on the CPU. The data does not include per-frame GPU clocks, but the FPS patterns imply the GPU is not running at its boost ceiling consistently. The fact that 4K Ultra (392.0 FPS) is only 10.4% behind 1080p Low (540.6 FPS) reinforces that the GPU has massive untapped reserves.

The GPU's benchmark scores contextualize its capability. It scores 38,194 in Passmark G3D and 31,7684 in Geekbench OpenCL, placing it in the 91st percentile of all GPUs. Its nearest rivals in the database are professional/workstation parts like the Tesla T4 (0.4% lower score) and Tesla P40 (0.5% higher), which are not gaming cards. The RTX 4090 is simply brute-force overkill for Valorant, and the measured FPS are constrained by the CPU, not the GPU's rendering limits.

Measured FPS Breakdown

At 1080p, the averages are as follows: Low runs at 540.6 FPS, Medium at 456.0, High at 449.5, and Ultra at 440.8. The gap between Low and Medium is 84.6 FPS, which is the largest jump between adjacent settings. Moving from Medium to High costs only 6.5 FPS, and High to Ultra costs 8.7 FPS. This suggests that the Low preset removes a significant CPU-side workload (likely shadows and physics), while Medium through Ultra are progressively more GPU-bound but still not pushing the GPU hard.

At 1440p, the pattern is similar but slightly lower. Low delivers 510.8 FPS, Medium 437.0, High 428.8, and Ultra 417.7. The Low-to-Medium delta is 73.8 FPS, again the largest. The 1440p results are on average about 5-6% lower than 1080p, which is a minimal penalty for the resolution upgrade.

At 4K, Low produces 474.4 FPS, Medium 399.9, High 396.8, and Ultra 392.0. The Low-to-Medium delta here is 74.5 FPS. Notably, the gap between Medium and High at 4K is just 3.1 FPS, and High to Ultra is 4.8 FPS. This indicates that at 4K, the GPU is finally starting to contribute slightly to the bottleneck, but the CPU is still the dominant constraint. The 4K Ultra result of 392.0 FPS is still far above the refresh rate of any mainstream monitor, so the experience remains exceptionally smooth.

FAQ

Q: Is the RTX 4090 being fully utilized in Valorant with the Ryzen 7 5800XT?

A: No. The data shows that FPS only drops 12.2% from 1080p Low (540.6 FPS) to 4K Low (474.4 FPS), which is a hallmark of CPU-bound performance. The GPU has 24 GB of GDDR6X and 1.01 TB/s bandwidth, far more than Valorant requires, so it is likely idling relative to its 2520 MHz boost ceiling.

Q: What is the best resolution to play at for competitive advantage?

A: 4K Low at 474.4 FPS is the best measured result at high resolution, but 1080p Low at 540.6 FPS is the highest overall. Given the minimal FPS difference, players can choose any resolution without a competitive penalty, as even 4K Ultra holds 392.0 FPS.

Q: How does the 5800XT compare to its closest rivals in CPU benchmarks?

A: The 5800XT scores 29,774 on average, which is within 0.3% of the AMD Ryzen 7 6800HS (29,856) and the AMD Ryzen 7 7840H (29,868). It is 0.1% ahead of the Ryzen 7 PRO 5755GE (29,740) and 0.3% ahead of the AMD Ryzen AI 7 PRO 360 (29,682).

Q: Does the Low preset offer a meaningful FPS advantage over Ultra?

A: Yes, but the absolute difference is more significant than the relative one. At 1080p, Low (540.6 FPS) is 99.8 FPS ahead of Ultra (440.8 FPS). At 4K, the gap narrows to 82.4 FPS (474.4 vs 392.0). The Low preset removes CPU overhead, which is why the delta is larger than typical GPU-bound games.

Q: Is the combo rank of 402 out of 1650 surprising given the hardware?

A: It is slightly lower than expected for a top-tier GPU, but it reflects the CPU's single-thread score of 959 in 3DMark. Valorant favors high IPC, and the 5800XT's Zen 3 architecture, while strong, is outclassed by newer designs. The 86th percentile CPU ranking is good, but not elite.

Q: What is the memory bandwidth of the RTX 4090 and does it matter here?

A: The RTX 4090 has 1.01 TB/s of bandwidth on a 384-bit bus with 24 GB of GDDR6X. In Valorant, this bandwidth is not a limiting factor, as evidenced by the flat scaling curve. The bottleneck is the CPU's frame generation rate, not memory throughput.

Settings Recommendations

The measured data points to a clear recommendation: play on Low preset for the highest possible frame rate, regardless of resolution. At 1080p Low, the combo delivers 540.6 FPS, which is the single best result in the dataset. The Low preset is not just marginally faster; it is 84.6 FPS ahead of Medium at 1080p, which is a significant margin even if the absolute numbers are all high.

However, for players who prioritize visual clarity over raw FPS, the Ultra preset is still eminently playable. At 1080p Ultra, the combo maintains 440.8 FPS, and at 4K Ultra it holds 392.0 FPS. These numbers are all far above the 240 Hz or 360 Hz refresh rates of the fastest competitive monitors. The difference between Low and Ultra is 99.8 FPS at 1080p, but both are so high that the choice comes down to personal preference for visual fidelity versus maximum responsiveness.

For the best balance of experience, the data suggests that 1440p High (428.8 FPS) is a sweet spot. It delivers near-4K visual quality with an FPS figure that is only 4.6% behind the 1080p High result (449.5 FPS). The 1440p Ultra result (417.7 FPS) is also strong. Given that the CPU is the bottleneck, the settings recommendation is to avoid Medium and High, which offer no real benefit over Low or Ultra in terms of the CPU load pattern. Low is for esports purists, Ultra is for those who want the game to look its best, and Medium/High are the middle ground that offers no distinct advantage.

CPU Role

The AMD Ryzen 7 5800XT is the primary determinant of performance in this Valorant pairing. With 8 cores and 16 threads based on Zen 3 architecture, it has a base clock of 3.80 GHz and a boost clock of 4.80 GHz. The CPU's 32 MB of L3 cache and 51.2 GB/s of dual-channel DDR4 memory bandwidth are more than sufficient for Valorant's engine. The CPU's single-thread benchmark score of 959 in 3DMark is the key metric, as Valorant is highly sensitive to single-core IPC rather than multi-core throughput.

The CPU's multi-threaded scores are strong—23,794 in Cinebench R23 multi-core and 7,683 in 3DMark 16-thread—but these are largely irrelevant for Valorant. The game typically uses a handful of threads, and the 5800XT's single-thread performance is what caps the FPS. The fact that the combo ranks 402nd out of 1,650 suggests that CPUs with higher single-thread scores would rank higher, even with the same GPU.

The CPU's percentile ranking of 86th among all CPUs places it above average, but its nearest rivals are all within a 0.3% performance band. This indicates that the 5800XT is not uniquely positioned; it is in a dense cluster of CPUs with similar capabilities. The 4.80 GHz boost clock is decent, but newer architectures (like Zen 4 or Zen 5) can push higher IPC and clock speeds, which would directly translate to higher Valorant FPS. In this specific game, the CPU is the bottleneck, and the data confirms that upgrading the CPU would yield more FPS than upgrading the GPU, which is already at the top of its class.