Fivem

The Fivem combo tested here pairs a AMD Ryzen 7 7700, a desktop 8‑core/16‑thread processor built on Zen 4, with an NVIDIA GeForce RTX 5070 Ti, the Blackwell‑era GPU with 16 GB of GDDR7 memory. The dataset is a full 4‑resolution, 4‑preset sweep at 1080p, 1440p, and 4K, with the frame rate numbers showing a decisive GPU‑bound pattern at higher resolutions and a more CPU‑sensitive behavior at the low setting. This review discusses how the scaling works and what each component brings, then closes with settings and ranking.

Resolution Scaling

The measured FPS across the resolution ladder tells a consistent story: as resolution rises, average framerates drop in a steep, almost exponential pattern. At Low, the sequence is 314 (1080p), 207 (1440p), and 115 (4K). At Medium, it goes from 247 → 163 → 90. High start at 198, set to 130, and then 72 at 4K. Ultra begins with 154, 101, and then only 56 at 4K. In every case the 1080p figure is more than twice the 4K figure, even though the pixel count from 1080p to 4K is a factor of 4. That means the GPU is being kept well‑fed at 1080p, but at 4K the rendering unit is limiting.

The scaling from 1080p to 1440p is rather gentle for a 720p pixel jump: at Medium, the drop is from 247 to 163, a ~95% reduction (in pixel terms) while FPS only drops ~36%. At the 1440p→4K step, the frame rate falls by about 60% of its 1440p value. That indicates that the card’s compute units and its 16 GB GDDR7 buffer are the main constraint at 4K, while the CPU’s instruction throughput becomes the limiter at lower resolutions. Since the 7700 absolutely has enough single‑thread to push beyond 300 at 1080p Low, the measured data shows that the scaling is not perfectly linear with pixel count, which is typical for a title that uses a whole‑ALU path.

The key takeaway from the resolution scaling is that if you are targeting 4K you are essentially paying for every ROP and shading unit, because the visual overhead of Fivem’s detailed geometry is enormous at that resolution. The 5070 Ti at 4K Ultra only produces 56 average fps, but at 1080p Ultra it reaches 154. So the resolution cap is not far off the GPU’s memory bandwidth limit.

GPU Role

The RTX 5070 Ti comes with a very large spec sheet: 16 GB of GDDR7 memory on a 256‑bit bus, a bandwidth of 896.0 GB/s, a boost clock of 2452 MHz, and a base clock of 2295 MHz. It has 49,600 million transistors on a 5 nm TSMC process, which is one of the densest GPU designs the companion uses. On the Fivem measured data, this hardware actually converts:

• At 1080p Medium the GPU delivers 247 average fps, with minima at 210 and maxima of 285.
• At 1440p Medium the averages drop to 163, with a min of 138 and max of 187.
• At 4K Medium it averages 90 (min 77, max 104).

Those are the only rows with min/max recorded, and they show that the GPU can push the frame rate well above any common refresh rate at 1080p and 1440p. The 96 ROPs and 686.6 GTexel/s texture rate will be seriously pressed at 4K, where at Ultra the card is producing 56 fps. Since the GPU and CPU percentages: the RTX 5070 Ti is at the 86th percentile among all GPUs, while the CPU is at the 87th percentile. That means this unit is well inside the top 20% for both.

The memory subsystem is crucial here. At 4K Ultra the scene complexity and terrain streaming can fill a 6‑GB buffer in a handful of frames, but the 16 GB allotment plus the 896 GB/s bandwidth is more than enough to avoid compromising framing. The GPU’s 70 ray‑tracing cores are never mentioned in the test settings, so the measured FPS are purely raster. With DirectX 12 (12_2), the card’s performance is set to the highest tier.

The measured FPS in this game’s actual pixel output rises significantly at lower resolutions, which indicates the shader-heavy load of Fivem correct. The 5070’s Tensor Cores are not directly required, but the 280 TMU count is high enough to keep texture fetch and filter busy.

CPU Role

The AMD Ryzen 7 7700 delivers 8 cores and 16 threads on the Zen 4 ‘Raphael’ architecture. Its base clock is 3.80 GHz with a boost up to 5.30 GHz. It combines four complex units with high single‑thread performance; note the Geekbench single‑core 2525 and Cinebench R23 single‑core 1930. The CPU has a 32 MB shared L3 cache, which is very helpful for the game’s world streaming.

In Fivem, a multiplayer mod environment, the game logic and vehicle physics can be CPU‑loaded. The numbers show that at 1080p Low the frame rate is 314, which is much higher than many single‑player games would achieve at the same setting. This implies that the system has to overcome engine bottlenecks. The Ryzen 7 also uses 65 W of TDP, which is much lower than the GPU’s 300 W.

The CPU offers strong raw single‑thread and multithread performance: in Cinebench R23, multicore score is 18760 and single‑core is 1930, while Geekbench multicore is 15371 and single‑core 2525. That is a clear balance. The 8‑core / 16‑thread configuration is enough to feed the RTX 5070 Ti at 1080p but the drop to 115 fps at 1440p (from 207 fps) suggests the CPU is still able to keep up while the GPU resolution climbs.

From a scaling perspective, the CPU does not appear to be a limiting factor in the lower settings until you reach the extreme of 4K Low where the average is 115, which is more than 60% of the 1080p Low value. That pattern is a typical CPU‑partial‑limit: the core speeds become optimal around the 165‑238 area, yet at 4K the GPU cannot produce more than 115.

Measured FPS Breakdown

The full set of measured FPS:

1080p

• Low: 314 avg
• Medium: 247 avg (min 210, max 285)
• High: 198 avg
• Ultra: 154 avg

1440p

• Low: 207 avg
• Medium: 163 avg (min 138, max 187)
• High: 130 avg
• Ultra: 101 avg

4K

• Low: 115 avg
• Medium: 90 avg (min 77, max 104)
• High: 72 avg
• Ultra: 56 avg

The measured media rows are the only ones with min/max. For example, 1080p Medium has a minimum of 210 and a maximum of 285, which is a 75‑25 spread. At 1440p Medium the min is 138 and the max is 187, so the fluctuation is about 35% at the midpoint. 4K Medium min 77, max 104. This points to a relatively stabile frame delivery at these resolution/settings.

Traversing down the table, a pattern emerges: the High preset at 4K (72 avg) is worse than the Low preset at 1440p (207 avg), but that is a resolution difference is more than. For any given resolution, Ultra is the most taxing, but you still gain double digits. Like at 1080p Low, the 314 clip makes it border on 8ms.

These values likely include both CPU and GPU usage, but there is no separate data.

How This Combo Ranks

In the global dataset of tested combos, this pairing of the Ryzen 7 7700 and RTX 5070 Ti lands at rank 378 out of 3707. That places it in the top 10% of all tested combos, which is a solid position, especially given the age of the CPU. It garners percentile ranks of 87th among all CPUs and 86th among all GPUs, so both parts are above average.

The nearest rivals in the CPU list: AMD Ryzen AI 9 365 (avg 40048) at þ1.0% delta, Intel Core 5 221E (40144) at -0.2%, AMD Ryzen 9 270 (40246) at -0.4%, and Intel Core i9-13905H (40313) at -0.6%. These are very close—within 1% of the average.

For the GPU, the nearest rivals are Radeon RX Vega 48 (50001) at -0.1%, Intel Arc A550M (49737) at 0.4%, RX 6900 XT (50951) at -2.0%, and RX 6800 XT (48477) at 3.1%. The 5070 Ti’s average is 49957, so it is remarkably tied with the Vega 64’s aggregate, though the Vega’s FPS is lower in actual game data.

The combo rank implies that the Ryzen 7 7700 is not the main reason, because despite being a top 20% CPU, the rank 378/3707 means unseen combos with much lower PC performance have matched this. The individual high percentages translate to good but not elite pairing.

Settings Recommendations

The settings that make the most sense depend largely on the display’s refresh rate:

• For 1080p: Most of the settings are extremely playable. The Medium setting averages 247 fps with the 210 minimum, which is perfect for any 144Hz or 240Hz monitor. Ultra is still at 154 fps, so if you want maximum visual fidelity, Ultra is the solid choice. For esports play, Low at 314 or Medium is certainly worth the slight look reduction, but at 1080p the GPU is able to use the CPU’s headroom.

• For 1440p: This is the sweet spot. High gives 130 fps, which is well above a 120Hz panel and extremely close to 144Hz. Ultra averages 101, still above 100, but if you have a 165Hz screen, Medium at 163 is the right pick. The min for Medium is 138, so there is sometimes a drop, but the average is competent.

• For 4K: The 5070 Ti is strong enough to drive a playable frame rate at High or Medium. Ultra only shows 56 fps, which is not comfortable for competitive play. Medium averages 90 (min 77, max 104) - that is more like a high‑end console experience. High gives 72, which is still near the 60 marker. Low at 115 offers the best guaranteed smoothness, without overly sacrificing clarity on 4K screens.

Recommendation: If you want a single preset across all resolutions, Medium is the best balance. It maintains a 90 average at 4K, 163 at 1440p, and 247 at 1080p, with min frames that stay believable. But if you have a 240Hz monitor, use Medium at 1080p; for a 1440p 240Hz display, Medium can exceed that; for 4K, you will need Low or High.

The component mix also works: the CPU is not a concern at these settings, as the FPS scaling is more GPU‑bound. With this pairing, if you want a 60 fps experience at 4K, it can handle High, but not Ultra.