Overpower

Overpower is an Action title from August 2023 that stresses both processor and graphics heavily. This pairing of an AMD Ryzen 5 3600 with an Intel Arc B570 produces a mid-pack combo, ranking 2373rd out of 2980 tested configurations. The data shows a system where the GPU is the dominant factor at higher resolutions, but the CPU's six-core, twelve-thread Zen 2 architecture still plays a decisive role at 1080p. The benchmark results indicate a playable experience across the board, but the settings you choose will determine whether you are GPU-bound or CPU-bound.

Resolution Scaling

The FPS drop from 1080p to 4K is steep, revealing a system that is heavily GPU-limited at higher resolutions. At Low settings, the average frame rate falls from 110 FPS at 1920x1080 to 84 FPS at 3840x2160, a 26 FPS reduction. This 23.6% performance loss when moving to 4K clearly shows the Intel Arc B570 is the primary bottleneck at lower graphical presets, as the CPU has enough headroom to feed the GPU at 1080p but the GPU's rendering workload at 4K becomes the limiting factor.

The scaling becomes even more pronounced when you look at the Medium preset. Here, the average FPS drops from 69 at 1080p to 62 at 1440p, and then further down to 52 at 4K. The gap between 1440p and 4K is 10 FPS, which is a larger absolute drop than the 7 FPS difference between 1080p and 1440p. This non-linear scaling is characteristic of a GPU that is hitting its memory bandwidth or shading limits as pixel count increases. The Arc B570's 380.0 GB/s bandwidth and 11.52 TFLOPS of FP32 performance are simply being saturated by the 8.3 million pixels of 4K.

At High settings, the pattern continues, with averages of 64 FPS at 1080p, 57 FPS at 1440p, and 49 FPS at 4K. The relative consistency of the drop (7-8 FPS per step) suggests that at this preset, the balance between CPU and GPU load is more even, but the GPU still leads the constraint. The Ultra preset shows the lowest absolute numbers, with 53 FPS at 1080p, 47 FPS at 1440p, and 40 FPS at 4K. The 13 FPS difference between 1080p and 4K at Ultra is the smallest absolute drop of any preset, which indicates the CPU is starting to become a more significant factor at this quality level, preventing the GPU from stretching its legs even at 1080p.

The data suggests that for a smooth 60 FPS experience, you need to be at 1080p with Medium settings (69 FPS) or better. At 1440p, only the Low preset (98 FPS) provides a comfortable margin above 60 FPS. The 4K resolution is only viable with the Low preset, which delivers 84 FPS, but the Medium (52 FPS) and High (49 FPS) presets fall below the 60 FPS threshold. The Ultra preset is not recommended at any resolution for a high-refresh experience, as it only reaches 53 FPS at 1080p.

How This Combo Ranks

With a combo rank of 2373 out of 2980, this pairing sits in the 79.6th percentile of all tested combinations in Overpower. This is a solidly mid-pack result, meaning that roughly 4 out of 5 systems tested will outperform this configuration. The ranking reflects a balance where neither component is a top-tier performer, but both are competent enough to avoid being a catastrophic bottleneck. The Ryzen 5 3600's percentileVsAllCpus of 71 and the Arc B570's percentileVsAllGpus of 65 indicate that both components are above average in their respective categories, but the combination does not boost to a high-performance tier.

The CPU's nearest rivals show how tight the competition is in this performance bracket. The Ryzen 5 3600's average benchmark score of 17035 is a mere 0.2% ahead of the Intel Core i7-1260P (17007) and 0.2% behind the Intel Core i5-11400 (17067). This places the Ryzen 5 3600 in a three-way tie with these two very different processors, one being a low-power mobile chip and the other a desktop part. The deltaPct values are so small that in real-world gaming, these CPUs are effectively interchangeable, meaning the GPU choice and game optimization matter far more than swapping between these three chips.

On the GPU side, the Arc B570's average benchmark score of 20556 puts it 0.1% ahead of the NVIDIA GeForce RTX 3070 Mobile (20534) and 0.1% behind the Intel Arc A750 (20582). The difference between the B570 and its direct predecessor, the A750, is negligible, suggesting that the B570 is a refined version of the same performance tier rather than a significant step forward. The closeness of these scores means that in Overpower, the specific graphics card model matters less than the driver version and in-game settings. The combo rank of 2373 reflects this parity, as the system lands in a crowded field where many different CPU-GPU pairings produce nearly identical results.

GPU Role

The Intel Arc B570 is built on the Xe2-HPG architecture with a BMG-G21 chip on TSMC's 5 nm process. It has 2304 shading units, 144 texture mapping units, and 80 raster output pipelines. The 18 ray tracing cores provide hardware acceleration for DirectX 12 Ultimate features. The card's base and boost clock are both locked at 2500 MHz, which is a modest clock speed for a modern GPU, but the 10 GB of GDDR6 memory on a 160-bit bus provides 380.0 GB/s of bandwidth. This memory configuration is the key to understanding the 4K performance, as the bandwidth is sufficient for 1080p and 1440p but becomes a limiting factor at 4K.

The GPU's FP32 performance of 11.52 TFLOPS is the raw compute throughput that Overpower's shaders will use. At 1080p, this compute power is more than enough to maintain high frame rates, as evidenced by the 110 FPS average at Low settings. The texture rate of 360.0 GTexel/s and pixel rate of 200.0 GPixel/s are also important for this action game, which likely uses high-resolution textures and particle effects. The B570's 80 ROPs are responsible for the final pixel output, and at 4K, these ROPs are processing four times the pixels of 1080p, which contributes to the significant FPS drop at that resolution.

The VRAM capacity of 10 GB is not a constraint at any of the tested settings in this game, as the measured FPS does not show a sudden cliff that would indicate memory swapping. However, the 160-bit memory bus is a potential bottleneck. The 380.0 GB/s bandwidth is the same as many 8 GB cards from the previous generation, and the data shows that the performance gap between 1440p and 4K is larger than the gap between 1080p and 1440p, which is a classic sign of memory bandwidth saturation. The card's suggested PSU is 450 W and it draws 150 W TDP, which is reasonable for a dual-slot card with a single 8-pin power connector.

Measured FPS Breakdown

At 1920x1080, the Low preset delivers the highest average FPS at 110, with no minimum or maximum recorded. The Medium preset averages 69 FPS with a minimum of 59 and maximum of 79, indicating a stable frame time distribution. High settings average 64 FPS, and Ultra drops to 53 FPS. The 1080p results show a clear hierarchy where each preset step costs roughly 5-6 FPS, but the jump from Low to Medium is the largest at 41 FPS, suggesting that Low is a special case that heavily reduces CPU load.

At 2560x1440, the Low preset averages 98 FPS, which is only 12 FPS lower than at 1080p. This suggests that at Low settings, the CPU is the primary limiter, as the GPU has enough headroom to handle the increased pixel count without significant performance loss. The Medium preset averages 62 FPS with a minimum of 52 and maximum of 71, showing a wider variance than at 1080p. High settings average 57 FPS, and Ultra averages 47 FPS. The drop from Low to Medium at 1440p is 36 FPS, which is a much larger gap than the 14 FPS difference between Medium and High, indicating that the Medium preset introduces a significant GPU workload increase.

At 3840x2160, the Low preset averages 84 FPS, which is still a playable frame rate. The Medium preset averages 52 FPS with a minimum of 44 and maximum of 60, meaning the game will dip below 45 FPS in demanding scenes. High settings average 49 FPS, and Ultra averages 40 FPS. The 4K results show that the GPU is the dominant bottleneck here, as even the Low preset cannot maintain the 110 FPS seen at 1080p. The difference between 1440p and 4K at Low settings is 14 FPS, which is a larger drop than the 12 FPS difference between 1080p and 1440p, confirming the GPU scaling trend.

Settings Recommendations

The measured data suggests that the best experience for a 60 FPS target is the Medium preset at 1920x1080, which delivers an average of 69 FPS with a minimum of 59 FPS. This provides a comfortable margin above the 60 FPS threshold and ensures that most scenes will run smoothly. The Medium preset also shows the most consistent frame pacing, with a narrow 20 FPS range between minimum and maximum. If you have a 144 Hz monitor, the Low preset at 1080p (110 FPS) is the only option that comes close to maximizing refresh rate, but the visual quality sacrifice is significant.

For 1440p gaming, the Low preset (98 FPS) is the only option that maintains a high frame rate. The Medium preset (62 FPS) is playable but will occasionally dip to 52 FPS, which may be noticeable in fast action scenes. If you prefer visual quality over frame rate, the High preset at 1440p (57 FPS) is a viable choice for a 60 Hz display, but you will experience occasional stutters during intense combat. The Ultra preset at 1440p (47 FPS) is not recommended for competitive play but is acceptable for slower-paced exploration.

At 4K, the Low preset (84 FPS) is the only setting that provides a smooth experience. The Medium preset (52 FPS) is borderline, with a minimum of 44 FPS that will be noticeable in action sequences. The High and Ultra presets at 4K (49 and 40 FPS respectively) are only suitable for screenshots or very slow gameplay. The data shows that this combo is not ideal for 4K gaming at high quality, and you should either drop to 1440p or accept the Low preset at 4K. For the best balance of visual fidelity and performance, the Medium preset at 1080p is the recommended starting point.

FAQ

Q: Can this system run Overpower at 60 FPS on a 4K monitor?

A: Yes, but only at the Low preset, which averages 84 FPS. The Medium preset at 4K averages 52 FPS with a minimum of 44, so it will not maintain a consistent 60 FPS. The High and Ultra presets at 4K average 49 and 40 FPS respectively, which is below the 60 FPS threshold.

Q: Is the CPU or GPU the main bottleneck in this game?

A: The GPU is the primary bottleneck at 4K and at higher quality presets. The data shows a 26 FPS drop from 1080p to 4K at Low settings, which indicates the Arc B570 is struggling with pixel throughput. At 1080p Low settings, the CPU becomes more relevant, as the GPU has headroom and the Ryzen 5 3600's single-thread performance of 696 in 3dmark limits the frame rate.

Q: How does this combo compare to a system with an Intel Core i5-11400?

A: The Ryzen 5 3600's average benchmark score of 17035 is 0.2% lower than the Intel Core i5-11400's score of 17067. This difference is negligible, and in Overpower, the two CPUs would perform nearly identically. The GPU choice would have a much larger impact on frame rates.

Q: What is the best preset for a 144 Hz 1080p monitor?

A: The Low preset at 1080p is the only option that gets close to 144 Hz, averaging 110 FPS. The Medium preset averages 69 FPS, which is below 144 Hz. If you want to use a high refresh rate monitor, the Low preset is the only viable option, but you will sacrifice significant visual quality.

Q: Does the 10 GB of VRAM on the Arc B570 cause any issues in Overpower?

A: The measured FPS data does not show any sudden performance cliffs that would indicate VRAM capacity problems at the tested settings. The 10 GB capacity is sufficient for this game at 4K Ultra, as the average FPS of 40 is consistent with the GPU's compute limitations rather than memory swapping. The 380.0 GB/s bandwidth is more of a constraint than the capacity.

Q: Is this combo better than a system with an Intel Arc A750?

A: The Arc B570's average benchmark score of 20556 is 0.1% lower than the Arc A750's 20582. This means the two GPUs are effectively tied in performance. In Overpower, the measured FPS would be nearly identical between the two cards, and any difference would be within the margin of error.