Valorant

# Valorant with Intel Core i5-10400F + NVIDIA GeForce RTX 5090

The combination of a 6-core Intel Core i5-10400F (Comet Lake, 14 nm, 12 threads, 2.90 GHz base, 4.30 GHz boost) with the flagship NVIDIA GeForce RTX 5090 (Blackwell 2.0, 5 nm, 92,200 million transistors, 32 GB GDDR7) produces an unusual pairing in a game like Valorant, a 2020-released FPS with modest visual demands. The measured data reveals a system that is heavily front-end limited at lower resolutions, with the CPU's single-thread performance (3dmark_single_thread score of 688, passmark_single_thread of 2541) becoming the primary constraint as frame rates climb past 300 FPS. At 4K, the GPU begins to show its influence, but the overall picture is one of a bottlenecked high-refresh experience.

Resolution Scaling

The FPS progression from 1080p to 4K shows a clear, consistent pattern: frame rates drop as resolution increases, but the magnitude of the drop is modest for a GPU with the RTX 5090's raw capabilities (passmark_g3d score of 39650, percentileVsAllGpus of 94). At Low settings, average FPS moves from 388.1 at 1920x1080, to 366.6 at 2560x1440 (a 5.5% reduction), and finally to 339.6 at 3840x2160 (a 12.5% reduction from 1080p). This relatively small penalty across a 4x pixel count increase indicates that the RTX 5090 is nowhere near its saturation point—even at 4K, the GPU's 104.8 TFLOPS FP32 throughput and 1.79 TB/s memory bandwidth are not being fully utilized.

The High settings row tells a similar story: 325.9 FPS at 1080p, 309.4 at 1440p (down 5.1%), and 287.3 at 4K (down 11.8%). Medium and Ultra settings follow the same proportional decline, with Ultra dropping from 316.1 to 304.3 to 279.6 across the three resolutions. The consistency of these deltas—roughly 5% from 1080p to 1440p and another 6-7% from 1440p to 4K—suggests that the limiting component is not the GPU's fill rate or pixel processing (176 ROPs, 423.6 GPixel/s pixel rate), but rather the CPU's ability to feed frames.

At 1080p Low, the system produces 388.1 FPS, which is 48.5 FPS higher than at 4K Low (339.6). If the GPU were the bottleneck, the 4K-to-1080p gap would be far larger given the 4x pixel count difference. The data indicates that the i5-10400F, with its 12 MB shared L3 cache and 42.7 GB/s memory bandwidth, is the constraining factor. The RTX 5090's base clock of 2017 MHz and boost clock of 2407 MHz are sufficient to handle Valorant's geometry, but the 6-core, 12-thread CPU (cinebench_r23_multicore score of 10297) becomes the ceiling for draw-call submission and game logic.

How This Combo Ranks

Within the database of 1,650 tested combos for Valorant, this pairing ranks 1,274th. That places it in the 22.8th percentile overall, meaning roughly 77% of all tested CPU-GPU combinations achieved higher average frame rates in this specific game. The rank is notably low for a system containing the RTX 5090, which itself ranks in the 94th percentile among all GPUs (avgBenchmarkScore of 84306). The disconnect between the GPU's high percentile and the combo's low rank underscores the severity of the CPU bottleneck.

To contextualize, the i5-10400F's nearest rivals in CPU benchmarks include the AMD EPYC 7552 (avgScore 14115, deltaPct 0.4 ahead), AMD Ryzen 3 7320C (avgScore 14277, deltaPct 0.8 behind), Intel Core i5-8500 (avgScore 14332, deltaPct 1.1 behind), and AMD Ryzen Embedded V2546 (avgScore 14336, deltaPct 1.2 behind). The i5-10400F's avgBenchmarkScore of 14168 places it in the 73rd percentile of all CPUs, which is respectable for general workloads but insufficient for pushing Valorant past the 400 FPS mark that faster processors can achieve. The CPU's passmark_single_thread score of 2541 is the key metric—Valorant's frame pacing relies heavily on single-core performance, and this score is modest by 2025 standards.

Among combos, the 1,274th rank out of 1,650 means the system sits in the bottom quarter. This is not a reflection of GPU failure; rather, it indicates that pairing a top-tier GPU with a mid-range 2020 CPU creates a mismatch in this game. The 3dmark_2_threads score of 1350 and 3dmark_4_threads of 2560 further illustrate that the CPU's multi-threaded scaling (3dmark_max_threads of 4735, near the 16-thread score of 4748) does not translate to Valorant's engine, which likely uses only a few threads.

FAQ

Q: Why does the RTX 5090 only achieve 287 FPS at 4K High when it is such a powerful GPU?

A: The bottleneck is the Intel Core i5-10400F. The CPU's single-thread score (688 in 3dmark_single_thread, 2541 in passmark_single_thread) limits frame submission. The RTX 5090's passmark_g3d score of 39650 shows it can handle far more, but the 6-core Comet Lake processor cannot feed it fast enough in Valorant.

Q: Is there a big difference between Low and Ultra settings at the same resolution?

A: At 1080p, Low averages 388.1 FPS while Ultra averages 316.1 FPS—a 72 FPS gap. At 4K, the gap narrows to 60 FPS (339.6 Low vs 279.6 Ultra). The delta indicates that higher settings increase GPU load, but the CPU remains the primary limiter, so the setting-related FPS penalty is relatively small.

Q: How does this combo compare to other tested systems in Valorant?

A: It ranks 1,274th out of 1,650 combos, placing it in the bottom 23%. The RTX 5090's GPU percentile (94th) is far higher than the combo's rank, confirming that the CPU drags the overall performance down.

Q: What CPU benchmark scores are relevant for Valorant?

A: Single-thread metrics matter most. The i5-10400F scores 688 in 3dmark_single_thread, 1453 in cinebench_r23_singlecore, and 2541 in passmark_single_thread. These scores are in the mid-range, and Valorant's engine appears sensitive to them.

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

A: No. Valorant's visual complexity is low, and the measured FPS values show no memory-related stutters or limits. The 32 GB GDDR7 with 1.79 TB/s bandwidth is massively overprovisioned for this title.

Q: Can the system reach high refresh rates like 360 Hz?

A: At 1080p Low, the average is 388.1 FPS, which exceeds 360 Hz. At 1080p High, it drops to 325.9 FPS, below 360 Hz. For 1440p High, the average is 309.4 FPS, so 300 Hz monitors are feasible but 360 Hz is not consistently achievable.

Measured FPS Breakdown

The measured data provides 12 distinct rows across four settings presets (Low, Medium, High, Ultra) and three resolutions (1920x1080, 2560x1440, 3840x2160). All averages are above 279 FPS, confirming that the system never dips below high-refresh territory, but the absolute numbers reveal the CPU ceiling.

At 1920x1080: Low averages 388.1 FPS, Medium 332.1 FPS, High 325.9 FPS, and Ultra 316.1 FPS. The spread between Low and Ultra is 72 FPS, or 18.5% of the Low value. The Medium-to-High gap is small (6.2 FPS), suggesting that these two presets impose similar CPU workloads.

At 2560x1440: Low averages 366.6 FPS, Medium 312.2 FPS, High 309.4 FPS, and Ultra 304.3 FPS. The Low-to-Ultra spread narrows to 62.3 FPS (17.0%). The step from Medium to High is only 2.8 FPS, indicating diminishing returns from the GPU side.

At 3840x2160: Low averages 339.6 FPS, Medium 291.1 FPS, High 287.3 FPS, and Ultra 279.6 FPS. The spread is 60 FPS (17.7% of Low). The High-to-Ultra gap is 7.7 FPS, the smallest delta among all resolutions for those two presets.

Resolution scaling per setting: Low drops by 21.5 FPS from 1080p to 1440p, then another 27 FPS from 1440p to 4K. Medium drops by 19.9 FPS, then 21.1 FPS. High drops by 16.5 FPS, then 22.1 FPS. Ultra drops by 11.8 FPS, then 24.7 FPS. The pattern shows that the 1440p-to-4K transition costs more than the 1080p-to-1440p step for all settings, which aligns with the GPU taking on more work as pixel count quadruples.

GPU Role

The NVIDIA GeForce RTX 5090 brings extraordinary hardware to this pairing: 32 GB of GDDR7 memory on a 512-bit bus delivering 1.79 TB/s bandwidth, 21,760 shading units, 680 TMUs, 176 ROPs, and 170 RT cores. Its base clock is 2017 MHz with a boost of 2407 MHz. The FP32 throughput of 104.8 TFLOPS (1:1 with FP16) and texture rate of 1,636.8 GTexel/s are overkill for Valorant's stylized, low-poly environments.

In the measured results, the GPU's role is largely passive. At 4K Ultra, the average FPS of 279.6 is the lowest recorded, but even this figure is far above what the RTX 5090's pixel rate (423.6 GPixel/s) would suggest as a hard limit. The passmark_gpu_compute score of 26756 and geekbench_opencl score of 380114 indicate massive compute headroom that Valorant does not tap into. The GPU's 94th percentile ranking (avgBenchmarkScore 84306) stands in contrast to the combo's 22.8th percentile, reinforcing that the RTX 5090 is not the limiting factor.

The GPU's memory clocks (1750 MHz, 28 Gbps effective) and 1.79 TB/s bandwidth are irrelevant here—Valorant's frame buffers are tiny relative to this capacity. The RTX 5090's nearest rivals in GPU benchmarks (RTX 5090 D with avgScore 84241, deltaPct 0.1 ahead; RTX 5050 Mobile with 84171, deltaPct 0.2 behind) show that even minor GPU variations would produce nearly identical FPS in this game, because the CPU is the bottleneck. The 5 nm TSMC process and 92,200 million transistors are simply not exercised by this workload.

Settings Recommendations

Based on the measured rows, the optimal preset depends on the target refresh rate. For players with 360 Hz monitors, only 1080p Low (388.1 FPS) clears the bar. At 1440p, the highest achievable preset for 360 Hz would be a hypothetical between Low (366.6 FPS) and Medium (312.2 FPS), but since Medium falls short, Low is the only viable option at 1440p for 360 Hz.

For 240 Hz displays, the recommendations broaden significantly. At 1080p, all four presets exceed 240 FPS, with Ultra at 316.1 FPS providing the best visual quality without sacrificing the target. At 1440p, all presets also clear 240 FPS—Ultra averages 304.3 FPS—so the highest visual fidelity is attainable. At 4K, Low (339.6 FPS) and Medium (291.1 FPS) both exceed 240 FPS, but High (287.3 FPS) and Ultra (279.6 FPS) fall below that threshold.

For 144 Hz monitors, every measured configuration is sufficient. The choice between Low and Ultra becomes purely aesthetic, with the 72 FPS penalty at 1080p (388.1 to 316.1) and 60 FPS at 4K (339.6 to 279.6) being acceptable trade-offs for higher visual fidelity.

The data suggests that Medium settings offer the best balance of stability and image quality for most users. At 1080p Medium, the 332.1 FPS average is within 14.4% of Low, while providing better visuals. At 1440p Medium, 312.2 FPS keeps the system above 300 FPS, which is a common high-refresh target. At 4K Medium, 291.1 FPS is the only preset (besides Low) that stays above 290 FPS, making it the recommended choice for 4K users who want more than the bare minimum settings. Ultra provides marginal visual gains over High (the delta is 9.8 FPS at 1080p, 5.1 FPS at 1440p, and 7.7 FPS at 4K) but at a consistent cost of roughly 5-8% performance, making it the least efficient preset for competitive play.